The Consequences of Providing and Refusing Refuge
Center for Tankship Excellence, USA, [email protected]
these events occurred, and then examine the follow-ing events to see how these decisions worked out.
This paper examines all the coastal state refusals andprovisions of refuge of stricken vessels in the Cen-ter for Tankship Excellence Tanker and Bulk Car-
rier Casualty database. According to this set of data,there is only one case where provision of refuge re-
Table lists the coastal state refusals currently in
sulted in a sizable (2000 ton) spill to the refuge
the CDB. We cannot of course claim that this list is
provider. On ther other hand, we have two cases
completeIt is simply the ones the CTX’s knows
where failure to provide refuge turned smallish to
about. The spill volumes shown are Known Volume,
moderate spills into two gargantuan spills totalling
In the case of the Castor, we know there was some
160,000 tons. In both cases, almost all the oil came
spillage, but we don’t know how much. Let’s see how
ashore on the refusing country’s coast.
identified at least 10 casualties in which provisionof refuge very likely prevented a 200,000 ton plus
spill. Most of the spillage would have come ashore onthe refuge provider. It appears that enlightened self-
interest can be a strong reason for providing refuge.
fully loaded 110,000 dwt tanker touched bottom onan uncharted ledge entering La Coru˜
tent of the damage is unclear, but at least the FPtank, one center tank, and the starboard bunker tankwas breached. It is probable all the center cargo
Tanker; Coastal State; Port State; Refuge;
tanks were holed on the flat bottom. But the crewtransferred fuel from holed bunker tank, and hy-
drostatic balance was quickly achieved in the cargotanks. The engine room and pump room were oper-
The purpose of this note is not to discuss the le-
ational. Weather was calm and clear. At that point,
gality or morality of providing or refusing refuge
no further significant spillage was possible without
to distressed vessels, but rather to examine the ac-
another screw up. The ship could easily have been
brought into sheltered waters, ballasted down, and
cellence (CTX) maintains a tanker and bulk car-
then lightered using salvage pumps for the breached
rier casualty data base (CDB). This is a public,
freely available, event centric data base.
However, the Port Captain immediately ordered
Urquiola 200 miles offshore without even the most
is described as a sequence of events, and each
cursory survey.. He explicitly required the ship to
event can be described in considerable detail. See
go out the same channel she had come in, despite
the fact that a deeper channel was available. At this
point, the ship’s draft forward was 1.5 meter more
Among the possible events are Coastal State Re-
than it was on its way in due to the flooding for-
fuses Refuge and Coastal State Provides Refuge.
ward, and the tide was lower. The ship stranded in
In this paper, we pick out the casualties in which
almost the same place she touched earlier. This time
1 Conspicuously missing from this list is the While it is true the Donges harbor master told the ship she could not
enter the Loire if she was leaking, the ship continued on a basically eastward course, until she broke up. The French may havebeen about to refuse refuge; but, if so, they never got the chance.
2 If you are viewing this paper on line, the ship names in Small Caps are links to the casualty’s precis file in the CTX
database. There you will find descriptions of the casualty drawn from various sources and/or links to such descriptions.
she hit hard with massive venting forward. She was
certainly did not cause any additional pollution, and
abandoned by all but the pilot and the master. Two
moved the pollution away from shore.
hours later the non-inerted tanker exploded, killingthe captain, and creating one of the largest spills inhistory. In all some 110,000 m3 of oil was spilled into
the ocean or burned into the atmosphere. Almost all
The was a bulk carrier in very poor condi-
tion. Crossing the Indian Ocean she suffered floodingin two of her seven holds and diverted to Mauritius.
She was refused refuge, escorted offshore, the crewtaken off, and allowed to sink.
Given the poor condition of the ship and the lim-
ited salvage resources available, it is not clear that
tained 30,000 ton tanker went aground at low speed
the ship could have been saved, in which case the
on Georges Bank about 29 miles from Nantucket.
coastal state’s action moved the bunker spill offshore
She was 24 miles off course. The ship was not badly
and prevented a possible blockage of the harbor.
damaged at the time. The US Coast Guard refusedthe master’s request to jettison 6,000 tons of fuel oilto refloat the ship. Instead the USCG invoked the
Intervention Convention — the first time this hadbeen done — and took charge of the salvage opera-
On the evening of 15th of February, 1996, the
tions. These operations failed. A week later the ship
broke up spilling 29,000 m3 of heavy fuel oil. The
Haven with 130,000 tons of light crude on board. The
winds and current kept the oil offshore.
pilot misjudged the tidal set and the ship grounded
In this case, no shoreline was impacted by the
on her starboard side, and then floated free, drift-
spill, but it is quite possible that the coastal state’s
ing into deeper water. 7 of her 14 cargo tank were
actions injected an additional 23,000 tons of oil into
damaged, but all the damage was very low. After los-
ing about 5000 tons, hydrostatic balance was quicklyestablished, and spillage stopped. The flooding ofthree ballast tanks further improved the hydrostatic
balance. At that point she could have been brought
into sheltered waters, with little or no additional
fered a massive structural failure just off the north-
west corner of Spain. She quickly lost 50,000 tons of
However, the Harbor Master effectively refused
her 208,000 ton cargo; but then the situation stabi-
entry, the weather worsened (as had been predicted),
lized, and salvors were able to take her under tow.
the tugs lost control of the ship, and over the next
Spain, Portugal, France and the U.K are known to
six days the ship re-grounded and re-floated at least
have refused refuge. The Andros Patria was towed
four more times. The final grounding was in very
to an area 200 mile south ofthe Azores, and eventu-
shallow water at the peak of the tide. Most of the
ally completed lightering north of the Canaries, as
total of the 72,000 tons spilled was lost on the sub-
the salvors moved ever southward in an attempt to
sequent ebb tide. Almost all this oil came ashore on
the Welsh coast. The Harbor Master’s policy turned
In this case, the ship’s survival indicates she could
a 5,000 ton spill into a 72,000 ton spill.
have been brought into sheltered waters where theprobability of a successful lightering would have been
a lot higher. But fortunately little if any additionalpollution resulted from the four refusals,
The is an example of a rare, and ratherpuzzling refusal of refuge by South Africa. The Trea-
sure was a Capesize bulk carrier in lousy conditionwhich experienced flooding in her Number 4 hold.
The loaded with 52,500 tons of Iranian crude
She sought refuge in Cape Town, but was denied
from Iran to Aden suffered an engine room fire off
and ordered 50 miles off shore. However, less than 10
the coast of Oman, cause unknown. The crew aban-
miles offshore she sunk near Robbins Island, a pen-
doned, the fire spread to the cargo, but salvors were
guin rookery. The crew was rescued but the resulting
able to take the ship under tow. Oman refused per-
1300 ton bunker spill oiled thousands of penguins.
mission to enter the Straits of Hormuz and the ship
The key issue here is could the ship have been
was towed to deep water with the intention of sinking
saved. Cape Town is well equipped with salvage re-
her. However, before that happened the ship suffered
sources. However, the fact that the ship sank so
quickly after being denied suggests that she might
In this case, the inability to get the fires under
not have been. In any event, the coastal state gained
control indicates the coastal state’s decision almost
tow by the twin screw, 13,500 hp tug Smitwijs Sin-gapore. The Japanese refused refuge. The ship was
towed to Taiwan which also refused refuge. Eventu-
of gasoline. When she attempted to enter New York
ally, the ship was lightered in open waters and towed
in October, 2000, the USCG found gasoline leaking
into at least two double bottom tanks and turnedher away. The ship diverted to Point Tupper for re-
Coastal state authorities appear to be unaware
pairs. Not clear what would have happened if the
of the fact that once a single screw ship is under the
Canadians — perhaps unwittingly — had not taken
control of a well-found multi-screw tug, she is less
likely to have a problem entering port than whenunder her own power.
The fully loaded VLCC boundfor Come-by-Chance in Newfoundland developed
hull cracks low in 1S. The ship transferred enough oilout of the tank to establish hydrostatic balance and
requested refuge. The Canadians initially refused
in the Bay of Biscay suffered a side shell failure in
and then several days later said the ship could come
ballast tank, 3S, in heavy weather. 3S, originally
in if 1S were completely emptied of cargo, which was
a cargo tank, was converted to a segregated ballast
both unnecessary and probably impossible without
tank without coating it. CTX would be surprised
illegally putting cargo into ballast tanks.pages
if it were not badly wasted. The flooding also ex-
507-508] The damaged ship diverted 3000 miles to
tended into 2AS, an original ballast tank. The ship
the Caymans where she was lightered.
took on a 24 degree list, taking the main engine off-
The Canadians took a big, unnecessary risk of a
line. The list was so high that cargo was leaking
from the tank lids and butterworth openings. Thelist was corrected by counter-flooding 2AP and 3P,
but this put the sagging moment 50% over the legallimit. However, the ship still had sufficient strength
to withstand the stresses if there were no substantial
with gasoline suffered a massive main deck crack
wastage. The ship asked for refuge, but this was not
in the western Med on New Year’s Eve 2001. She
only denied by Spain and Portugal, but the ship was
was refused refuge by just about every country in
the western half of the Mediterranean in part be-
It is not clear exactly when the ship started leak-
cause of the volatile nature of the cargo. She was
ing again. Some Spanish sources claim that the cargo
towed around the Med for 40 days before finally be-
tanks were holed early in the process. This appears
ing lightered in open waters well off MaltaSub-
to be based on the areal extent of the original slick
sequent inspection of the ship revealed the steel in
combined with an unrealistically high estimate of the
the area of the deck failure was in horrible condition
spill thickness. Other sources claim she did not start
despite the fact that she had successfully passed all
leaking again until well into the tow off-shore.
In this case, none of the coastal states suffered
In any event six days into the tow, the ship failed
as a result of the refusal. But the safety argument is
in sag and sank creating a 72,000 ton spill. Oil seep-
bogus. It is true that there was an obvious argument
ing from his wreck will probably contaminate Euro-
for not bringing the ship into a port, but refuge in
semi-protected waters several miles offshore, prefer-
Unlike most of the casualties in this paper, this
ably downwind, could have been provided without
is not an easy call, even in hindsight. Given that the
any risk to people ashore. In any event, the suc-
ship survived six days, she almost certainly would
cessful lightering of the ship in totally unprotected
have survived the short tow to sheltered waters. On
water pretty much proves it could have been done in
the other hand, few would predict that a ship, whose
steel was probably in bad condition and had alreadystarted to fail, could handled stresses 50% over legal
limits for any length of time. Forcing such a shipoffshore, practically guaranteed a 72,000 ton spill;
but one whose impact would be spread over a wide
loaded with 249,000 tons of crude when it suffered
area. The courageous decision would have been to
a massive crankshaft failure off Japan her destina-
bring the ship in, but it would have been risky for
tion on 2nd May 2002. The ship was taken under
3 Transport Canada spins this incident slightly differently, claiming the owner diverted to the Caribbean “due to inclement
weather concerns”. This is highly implausible.
4 She was offer refuge by Cyprus, but did not take advantage of this, presumably because of the distance.
bow damage was allowed into Port Elizabeth, appar-ently without further spillage.
To balance the books a little, Table lists the ca-sualties in which a coastal state’s provision of refuge
We know this list is woefully incomplete.
the case of South Africa alone, there are something
had a major side shell failure in eastern Indian Ocean
like 100 tanker casualties, Table between 1966
in May, 1975. and lost 15,000 tons of crude from
and 1984 which can be construed to be provision of
tank wastage. She was allowed into the Dampier
The problem is that in almost all these cases the
Archipelago and the cargo lightered with nil further
coastal state response was so successful in preventing
a spill, that the casualty received so little attentionthat we have insufficient data to enter it in the CDB.
Nonetheless the CDB does contain some instruc-
tive examples of a coast state providing refuge.
Perhaps the most dramatic provision of refuge in theCDB involved the
Pacific Colocotronis, loaded with 72,000 tons of
volatile Algerian crude for Wilhemshaven, was off
the Dutch coast on the 28th of September, 1975 when
which grounded rounding Cape Point in April, 1968.
she started leaking cargo in way of 3P. The weather
Clearly, she was not 12 miles offshore as required by
was good, but inspection by the Dutch Coast Guard
the legislation of the time. We don’t know why. Six
revealed the damage was much worse than the ship
of her 12 cargo tanks were breached. Thousands of
had reported with at least two very large side shell
seabirds were oiled. She did not strand and was al-
cracks. Colocotronis ships had a habit of getting into
lowed into False Bay. Sources vary, but the most
trouble. The CTX is confident that the root prob-
likely spillage was 4000 tons, which would have oc-
lem was a badly wasted port segregated ballast tank.
curred very quickly after the grounding.
We can also be confident that the starboard ballasttank was in no better condition. The weather was
The ship was pretty obviously in no condition to
forecast to worsen. The ship was in no condition to
proceed. We know very little about the damage; but
proceed. On 1st of October, Ijmuiden allowed the
there is a good chance that, if she had not found im-
still leaking ship into its harbor. In all the Pacific
mediate refuge, the ship would have sunk somewhere
Colocotronis spilled about 1,800 m3. But Ijmuiden’s
near the South African coast a la Treasure, releasing
courageous action almost certainly prevented a fur-
ther 85,000 m3 of toxic light crude spilling into theNorth Sea. Much of that oil would have ended up on
After ramming and sinking the Texanita in Au-gust, 1972 with a loss of 44 lives and 8,000 to
10,000 tons of oil, the loaded 100,000 ton tanker
Fully loaded VLCC adrift off South Africa after ma-
jor machinery failure. She was towed to Algoa Bay
lightered, and dry docked. She suffered massive dam-
and lightered. 225,000 ton spill averted.
age to her bow, but apparently the damage did notextend into the cargo area. There is no record of anyfurther spillage.
On 16 December 1977, two 325,000 ton tankers the
and collided 40 miles off SouthAfrica. The Venoil was loaded with 312,000 tons of
Repeat of Oswego Guardian two years later. Loaded
crude. Both ships were holed, spilled oil, caught fire
and were abandoned. Both ships were taken under
of Tekton which eventually sank. Ship with massive
tow, allowed to drift in the Agulhas Current until the
5 The is not on this list. This is a judgement call. After the Kirki lost her forepeak tank 23 miles off the southwest
coast of Australia, the Australian required the ship to be towed 700 miles north to an area north of the Dampier Archipelagothat could hardly be called a place of refuge. During the tow, the ship lost another 10,000 tons of cargo. On the other hand, theship was in horrible condition, 64,500 tons was successfully lightered, and the Australians could have towed the Kirki straightoffshore and sunk her.
Nor have we included the After the Tanio broke in two off the north coast of Brittany, the stern section was towed
aroung the English Channel for several days before before being allowed into Le Havre.
If your definition of provision of refuge is broad enough to include these two casualties, the basic conclusions of this paper
6 Table was extracted from Brookes . The full list is at the
fires were out and they had stopped leaking, and then
towed into Algoa Bay, where the remaining 289,000
tons of oil on the Venoil was lightered.
off Algeciras, where she was anchoreduntil being towed to Gibraltar on Jan-
uary 28, 1980. After being declared aconstructive total loss, she was towed to
wasted forepeak tank simply fell off in bad weather. Ship was only six years old. There was some cargo
The similarities to the Erika and Prestige are ob-
leakage. She apparently proceeded to St Helena Bay
vious. But the coastal state reaction was entirely
(some say False Bay) under her own power and was
different and the resulting spill, according to Etkin,
lightered. Another 220,000 ton spill averted.
was only 300 m3. It may have been a bit more, butthe fact remains courageous coastal state reactionturned an Erika into an event no one remembers.
ingly bad ship performance and remarkably profes-sional coastal state response. The Christos Bitas,
loaded with 35,000 tons of crude, was proceeding
loaded VLCC suffering serious damage to a badly
north in the Irish Sea, bound for Belfast. One radar
corroded forepeak tank off South Africa. In this case,
had not been working for two years, the RDF had
the damaged involved the forward bunker tank. De-
not been calibrated for four years.On entering
spite the fact that the ship was leaking, the South
the Irish Sea, she lost her other radar. Dense fog,
Africans allowed the ship to be towed into Algoa Bay
but she proceeded blind, not even slowing down.
and lightered. The ship was in such bad condition,
She grounded on rocks off Milford Haven holing at
she was scrapped. It is quite possible a 250,000 ton
least half of her 12 cargo tanks. Three hours later
the Christos Bitas refloated herself, and the Masterblithely proceeded toward Belfast with the ship leak-
She was intercepted by the Royal Navy. With the
The a 110,000 ton tanker was fully loaded,
weather deteriorating, the Christos Bitas was towed
Venezuela to Rotterdam, when the side shell failed
toward the Irish coast where lightering could proceed
in way of the midship ballast tanks in heavy weather
in the lee of the land. In an difficult, complex ten day
off Portugal, resulting in an 8 m by 4 m hole. There
operation, the salvors managed to off-load almost all
was no spill, but the tank flooded, and the ship was
the cargo, before towing the ship out to deep water,
adrift for reasons unknown. She was towed to Sines,
and sinking her 300 miles off Ireland.
discharged, then towed to Setubal. The ship was in
This should have been a 35,000 ton spill in the
such bad condition she was scrapped. This is almost
Irish Sea. Thanks to the intelligent, courageous re-
certainly another case of segregated ballast tank cor-
sponse the actually spillage was about 4,000 tons.
It would not have taken much for the damage
to spread to the neighboring cargo tanks. The Por-
tugese response may have averted as much as a
because of the contrast in coast state performancewith the Prestige 22 years later. Here’s Hooke’s de-
last tank wastage. This loaded 57,000 ton tanker
suffered severe side shell damage off the Yucatan
in heavy weather; but the damage apparently did
not extend into the cargo tanks. She diverted to
Freeport, Bahamas discharged her cargo and was
105 miles north of Corunna in lat 45.05N,
long 07.50W on December 15, 1979. With oil seeping out and water enter-
ing in, the 37 crew requested immedi-ate assistance but, due to the atrocious
rier, was loaded with gas-oil Amsterdam to the US
ceeded to the area could get close enough
Gulf, when she encountered heavy weather in the
Bay of Biscay. She made it into La Coru˜
of February, 1984 where she was found to have a
was some spillage. Hold 8 was the aftmost hold. Ap-
warped deck and a badly damaged hull. The dam-
parently the damage, a 21 m by 10 m hole, involved
age was so bad that she was declared a CTL on the
the starboard fuel oil tank. The ship had 550 tons
of bunkers on board, probably half of which was in
Another precursor of the Erika and Prestige, but
this tank. The spill appears to have taken the form
thanks to the provision of refuge the lucky crew sur-
of a slow leak which was largely contained by booms.
vived and apparently there was no spill.
CTX has assigned a nominal spill volume of 50 m3to this casualty.
The Tribulus was alternate hold loaded with No.
8 empty. This means she was in serious danger of
sinking as many bulk carriers have done in this sit-
byshire and a deeply flawed vessel. On 18 November
uation. In fact, most of the crew was helicoptered
1986, this 169,000 ton OBO was loaded with iron
off the ship well offshore. The Irish decision to allow
ore when she experienced a major deck failure in
the Tribulus in probably saved the ship and most of
the North Atlantic off Ireland. She was allowed into
the bunkers at the cost of some inshore spillage.
Bantry Bay and received some repairs. But threedays later she managed to lose her anchor, and de-cided to head for open water. But then she lost her
steering, stranded near the entrance to the bay, andsplit in two forward of the bridge. The ship had 2000
tons of bunkers on board. There is no mention of
long history of problems. Here’s a first hand account
any recovery. Several local sources claim the wreck
of what happened to her in August, 1991.
leaked oil over a multi-year period. CTX is assuminga 2000 ton BFO spill. This is a rare case in which
the coastal state suffered a sizable spill as a result of
providing refuge. Of course, there’s a decent chance
south tip of Africa. We were fully loaded
the ship would have fetched up on the Irish coast if
In 2005, there was an attempt to salvage the
First we had a steering gear break down.
wreck and the iron ore. Dive groups bemoaned the
A non-return valve in the valve block be-
loss of a spectacular dive site and the damage to the
tween the two cylinders cracked, the pipe
marine life that flourished around the wreck.
connected could not take the pressure,and, all the oil in the system went on
the flooring. The rudder slammed fromside to side as the outside sea would drive
Quebec loaded when the master reported flooding
into double hull ballast tanks. The ship was diverted
to Halifax where divers found large cracks in the bilge
area. The 75,000 ton cargo of crude was lightered
and the vessel dry-docked. There was an attempt to
blame ice for the damage but the bilge on a loaded
75,000 ton tanker is 10 m below the waterline.
go dead slow astern for some 12 hours,and we went in a circle in the hurricane.
This probably saved us from ending upon the South African rocks.
The was a loaded 169,000 ton tanker whose
engine room flooded in heavy weather off Cape Town
in October 1988, cause unknown. She was towed into
”free” again. By then the seas had sub-
Table Bay, and then to Saldanha Bay and lightered.
sided and we got tug boat assistance.
She was in such bad condition, she was scrapped ina rising market.
When it was safe to go out on deck, wesaw a 340 m2 hole in ballast tank 5S. Theshell plate was nowhere to be found. And
yes, corrosion did play a major role in this
The was a 127,000 ton bulk carrier loadedwith iron ore, when she had a massive side shell fail-
Fortunately, the ballast tank damage did not ex-
ure in way of Hold 8 southwest of Ireland in heavy
tend into the neighboring cargo tanks. The South
Africans towed the Mimosa into Algoa Bay where
where she stayed for three months repairing. There
she was lightered averting a 350,000 ton spill.
loaded with crude when she lost her badly corroded
had a major machinery failure fully loaded off South
Forepeak tank off South Africa in heavy weather.
Africa. She was towed to False Bay, lightered, and
The tank simply fell off. The ship was towed to
Algoa Bay and lightered. A 250,000 ton spill was
averted, not that they had much choice.
59,000 tons of gasoline when she suffered a major
ternate hold loaded with iron ore when the bulkhead
machinery failure off South Africa. She was towed
between Hold 8 and Hold 9 collapsed off Norway.
Earlier the crew had gone to ITF complaining aboutthe condition of the structure, something crews only
do in the most extreme situations. The ship was ableto make it into Kristiansund.
The fully loaded 95,000 ton tanker wasabout to enter Sydney when she started leakingcargo. Divers discovered a 15 cm flat bottom crack
in 3C. They applied a magnetic patch, and she wasescorted into port.
The was a Capesize bulk carrier loaded withiron ore when she suffered extensive structural fail-ures off southern Brazil. She put into the Rio Plata.
On inspection, her steel was in such bad shape they
could not even safely discharge the cargo. She was
and the both loaded, and on nearly paral-
towed out to sea and sunk with the cargo still on
lel courses somehow managed to get into a collision
north of Port Said after leaving the Suez. The dou-ble sided Kestrel was holed in three starboard tanks,
spilled about 1000 m3 of cargo, took on a big list,which shut down her main engine. She was towed to
The another Capesize bulker, got a bit fur-
Cyprus and lightered in the lee of the island.
ther. She also was loaded with iron ore, Brazil to theFar East when she suffered a 47 meter long crack in
her side shell off Cape Town. She was allowed intoFalse Bay and repaired.
The part-loaded 300,000 ton tanker had a blackout in the Suez Canal, grounded, holingat least one ballast tank, but there was no spill. She
was towed to Cyprus and lightered in the lee of the
had suffered both war damage and bad maintenanceduring the Iran-Iraq War. In June, 1996 her badly
corroded Forepeak tank fell off 90 miles northwestof Cape Town. She was fully loaded. The forward
It will come as no surprise that in an issue like this
fuel oil tank was involved and there was some BFO
there are no hard and fast rules. But it is also sur-
spillage. The condition of the forepeak tank was so
prisingly clear that on the whole the countries that
bad, the decision was made to tow her stern first to
provided refuge have fared better, from a purely self-
False Bay, where she was lightered. A 300,000 ton
ish point of view, than the countries that refused
refuge. There is no case in our admittedly small sam-ple where a country suffered grievous harm from al-
lowing refuge. The worst that we can find is the 2000ton bunker spill from the Kowloon Bridge that might
In January, 1995, we renew acquaintances with our
have happened anyway if refuge had been refused.
old friend the This time the fully loaded
There are at least two cases, the Urquiola and the
357,000 ton ULCC suffers massive forepeak tank
Sea Empress, where the country that refused refuge
damage off Scotland. She was allowed into Lyme
did suffer massive damage from unnecessary 70,000
Bay, not the safest of places, and lightered. This
ton plus spills. And there are many cases where it
was a very risky decision on the part of the UK, but,
is quite likely that the country that provide refuge
given the alternatives, probably the right one.
would have suffered grievous harm if she had done
the opposite. As Table indicates, if South Africa
had not followed a policy of providing refuge, thereis good chance that she would have suffered a score
 ABS. Investigation into the damage sustained by
of 200,000 ton plus spills. Conversely, the only ma-
the m.v. castor on 30 december 2000, final report.
jor casualty in which you can plausibly argue that
Technical report, ABS, October 2001.
the refusing country came out ahead is the Prestige,
 Colyn Brookes. Marine casualty database south-
and even then the case is a weak one, and a 72,000
ern african coast. Technical report, NCS, 2008.
ton spill that turned out to be preventable was notprevented.
 R. Cahill. Strandings and their Causes. The Nau-
So why have refusals become the norm? The
major reason of course is the dichotomy betweenthe risk/reward for the refusing authority and the
 A. Chircop and O. Linden. PLaces of Refuge for
risk/reward for society as whole. In the past this
Ships. Martinus Nijhoff Publishers, 2005.
was balanced by the professionalism of the maritime
 A. Constantinou. Places of refuge, a myth or re-
authorities and their concern for crew safety. The
ality. Technical report, International Conference
authorities have become increasingly politicized and
on Martime safety and Protection, 2007.
the crews can always be taken off by helicopter.
But there is one technical thread in our refusals,
 Sindicato Libre de la Marina Mercante. Urquiola,
that perhaps we can do something about and that is
la verdad de una catastrofe. Technical report,
a lack of understanding of the efficacy of hydrostatic
balance on the part of coastal state authorities. Oncea tank reaches hydrostatic balance and stops leaking,
 Jack Devanney. The argument for twin screw
which normally is very quickly, especially in ground-
ings, that tank is not going to leak any more unless
the ship is stranded and the tide goes out or someone
 Norman Hooke. Maritime Casualties, 1963 to
starts discharging the intact tanks first. And if the
ship can be ballasted down, so much the betterIfthe refusing authorities understood hydrostatic bal-
 P. B. Marriott. Report on the grounding of the
ance, then there is a chance that some of the more
tanker sea empress at milford haven between 15
egregious mistakes, such as the the Canadians han-
and 21 feb 1996. Technical report, Marine Acci-
dling of the Eastern Power could be avoided.
7 Often the ship will ballast herself down by flooding ballast tanks. This happened in the case of both the Urquiola and the
Based on CTX Casualty Data Base as of 20090216
Based on CTX Casualty Data Base as of 20090216
Table 3: South Africa Provisions of Tanker Refuge
253,358 55 mi South East of Cape Point Rudder damaged, repaired
317,500 12 miles South of Cape Agulhas Power failure, towed to North of Robben Island
229,000 700 miles North East of Durban Engine room fire, towed to Cape Town
Hull damaged, cgo transhipped to Torill Knudsen in
75,000 180 miles South East of Durban Collision, proceeded to Durban
265,040 1500 miles North of Cape Town Engine room fire, repaired in Cape Town
213,373 4 mis South of Umhlanga Rocks Power failure, repaired
Hull damage, transhipped cargo to vessel Regina in
Propeller damaged, new blade cut in Cape Town
Caught fire, explosion. towed to Walvis Bay
Caught fire, explosion in the engine room,
Explosion in engine roome, towed to Cape Town
Oil leak, put to sea and repaired in Walvis Bay
Power failure, towed to Walvis Bay, then to Cape
Explosion in engine room, proceeded to Walvis Bay
Collision, caught fire. Proceeded to Cape Town
Power failure, towed to Algoa Bay, cgo transhipped
Power failure, proceeded to Saldanha Bay, cargo
Collision, towed to Mossel Bay, transhipped cgo to
Steering gear damaged, towed to Richards Bay
330,954 39 miles SW of Cape St Francis Collision, fire. Towed to Algoa Bay. Cargo
330,959 39 miles SW of Cape St Francis Collision, towed to Algoa Bay then to Cape Town
Steering gear damaged, repaired in False Bay
Steering gear damaged, repaired in Port Elizabeth
Hull damage, proceeded to Algoa Bay. Cargo
Hull damaged, proceeded to St Helena Bay. Cargo
173,715 20 miles West of Cape Agulhas Power failure, towed to False Bay
Climate change and its impact on health in Bangladesh Climate change affects human health both directly and indirectly. People are exposed This background paper was prepared for a workshop on Climate Change and Health in (temperature, precipitation, sea-level rise and Bangladesh, held on 19–20 November 2007 more frequent extreme events) and indirectly through changes in the quality of wa