Back to Casualty List | Search The Casualty Database
Precis File
SHIP NAME: Derbyshire KEY: NUM. ENTRIES: 9
source LINK
type D
volume
material
dead 44
link http://www.mv_derbyshire.org.uk

Agreed statement of facts.


source LINK
type L
volume
material
dead 44
link http://www.nautical-heritage.org.uk/derbyshire.html


source LINK
type L
volume
material
dead 44
link http://www.shipstructure.org/derby.shtml


source Faulkner, D., An Analytical Assessment of the Sinking of the M.V. Derbyshire, Royal Institution of Naval Architects, 2001
type L
volume
material
dead 44
link http://www.wemt.nl/RINA paper.pdf

The definitive study of this casualty. Faulkner concluded that easily the most likely cause of the sinking was failure of under-designed but legal No 1 hold hatch covers. But in 2003, Faulkner pointed out that the possibility of an overall hull girder collapse should be given more weight than he gave it in this paper. CTX feels that the 2003 paper is correct in noting out that the ship's structure was weak enough so that hull girder collapse was a real possibility in the conditions she encountered, but that the photographic evidence from the underwater surveys argues strongly that this did not in fact happen.


source LINK
type L
volume
material
dead
link http://www.nationalarchives.gov.uk/ero/browse.aspx?id=41&level=4

This is the official Assessors report published in March 1998. The site includes considerable underwater video, albeit at reduced resolution. CTX strongly recommends that anyone seriously interested in the Derbyshire study the pictures carefully.

By this time Professor Faulkner had been forced to resign as an Assessor (see CTX entry below). The two remaining Assessors, Robin Williams and Remeo Torchio, both naval architects, concluded that the Bosuns Store flooded due to an improperly secured hatch cover. Mr. Williams was particularly out-front in concluding that the crew was at fault. and suggesting that the crew left a manhole from the Bosuns Store to the partially filled Forward Fuel Oil tank open. a comment that showed complete ignorance of how crews work. The Forepeak tank flooded either thru damaged tank vents or maybe the crew left a manhole from the Bosuns Store to the Forepeak tank open. The latter is totally unsupported, extremely unlikely, and unwarranted. See Woinin entry below. The Assessors admit this process would have taken 12 hours or more. This reduced the forward freeboard by 2.4 m. For the Assessors this slow-filling was the Primary Cause or in their words "the initiating event".

The reduced freeboard forward combined with the weather generated a situation in which the No 1 hold hatch covers were over-stressed; they failed; 10,000 tons of water flooded in; then the No 2 hatch covers imploded, and this process continued bow to stern. To the Assessors the failure of the hatch covers was a secondary event. The Assessors's implication is that, if it had not been for the flooding of the spaces forward of No 1 cargo hold, possibly cause by crew error, she would have survived.

This slow filling of the forward spaces scenario has been critiqued by a number of commentators. See for example the Woinin paper in the next entry.

The CTX would add:

  1. The little hatch cover to the Bosuns Store was indeed horribly designed. You can see its location in this picture. It is the little rectangle on the centerline forward of the foremast. The ship had a design freeboard of only 6.6m and no raised forecastle. Submerging the bow was common place. Green water on top of the hatch cover, compresses the packing, releasing the toggles, however tightly screwed. But this was well-known by the crews. The Derbyshire's Bosuns Store had flooded just four months earlier. A similiar incident occured on the sister ship Furness Bridge in 1973. One attempt at a solution is to rig a cats cradle on the hatch cover. In fact the Derbyshire crew did just this. Remnants of the line from this attempt wrapped around the toggles show up clearly in some of the photos. And, in this case, the crew's efforts may have succeeded. The pictures. of this hatch indicate that the cover was torn off from impact on the aft coaming, possibly by the starboard windless which was ripped off its foundation. In any event, there is a more likely means of flooding the Bosuns Store which the Assessors pretty much ignore which is the mushroom ventilators which were ripped off.
  2. The Bosuns Store was about a 1000 m3 space. Flooding this space would decrease the forward draft by about 0.4 meters. By itself, this is almost a non-event as the earlier Bosuns Store floodings prove.
  3. The hull volume forward of the No 1 cargo hold is divided into four spaces: a cofferdam at Frame 339 just forward of the No 1 hold, a Forward FO tank forward of the cofferdam which extends up to the main deck, a Forepeak ballast tank forward of the Forward FO tank, and the Bosun's store above the Forepeak ballast tank. The flooding of the Bosuns Store prior to sinking is not improbable. Slow flooding of the Forepeak tank through its ventilators is possible but this would take a long time. But we can see from the underwater photos that the Forward FO tank and the cofferdam flooded very rapidly. See below. The Assessors ignore this evidence. Flooding of both the Bosuns Store and Forepeak tank would reduce the forward draft about 1.4 m, a reduction that a robust ship should have had no trouble in handling. In fact,there is evidence that the ship had a 1 to 1.5 m trim by the stern prior to any damage in which case this amount of flooding would merely bring the ship to even keel. The freeboard forward prior to damage was about 750 mm above the legal minimum.
  4. Lete's assume incorrectly that all the spaces forward of Frame 339 flooded slowly. Why did the ship not send out a distress signal? The Assessors are correct that a 2.4 m change in trim would have been impossible to ascertain visually at night in the middle of a big storm, when no one could go forward, and the only light switch was forward. However, it would clearly show up on the draft gauges. The Captain would have known immediately what that meant and at a minimum sent out a distress signal.
  5. The Assessors main argument is based on the fact that the bow. was relatively intact, while the cargo portion of the hull further aft had been shattered by a series of implosions. To the Assessors this is conclusive evidence that the bow was flooded prior to the sinking. This conclusion is unwarranted. With the exception of some of the bunker tanks, the stern structure aft of Frame 65 did not implode yet no one is suggesting it was flooded prior to sinking. Like the stern, the bow structure is a combination of quite strong, the shell, and quite weak, Forepeak tank top, and transverse bulkheads. It is quite possible for the weak portion of such structure to fail quickly and massively enough so that it floods prior to sinking to implosion depth. In fact, we know this is exactly what happened to the Forward FO tank and the cofferdam forward of the collision bulkhead. These spaces did not implode, they exploded; that is, they flooded from somewhere low, the water rushing in compressed the air in the top of these spaces, and blew the structure outward. The video evidence for this is conclusive, admitted by the Assessors, and then ignored. Faulkner thinks the cofferdam and Forward Fuel oil tank flooded from a large horizontal crack in the bulkhead at Frame 339 about 12 m above the baseline. Others have argued that this crack occurred when the bow hit the bottom. Faulkner has calculated the forces associated with this scenario and finds they are too small. The CTX's view on this is: we don't know. For oen thing we can't see the the lower 8 m of this bulkhead. If the cofferdam/forward FO tank did not flood from the No 1 cargo hold, then the most likely source is the FP tank. But this does not necessarily imply a slow-filling. It is only evidence that the Forepeak tank flooded prior to the Forward FO tank. Unlike the Forward FO tank and the cofferdam we cant see any of the weak surfaces of the FP tank so we don't know if it exploded in a similar manner.
  6. Some of the testimony indicates that the Forward FO tank contained 2000 to 2600 tons of bunkers. The reports are distressingly unclear on this one. If this was the case, flooding of this tank (slow or fast) would have had almost no effect on forward freeboard. It is possible that the Master transferred bunkers aft to increase the freeboard forward as he went into the storm. In which case slow-flooding if it had happened which it did not, would have had more impact, but only to bring the freeboard back down to that assumed by Faulkner and others.
  7. Finally, even if the Assessors incorrect hypothesis that all the spaces forward of Frame 339 filled slowly were correct, calling this filling the primary cause is very strange. A 2.4 meter trim by the bow should not have sunk a robust ship whose initial freeboard forward was 0.75 m above the legal minimum. The crucial event was the flooding of the No 1 hold. And for that we have very strong evidence that the cause was a hatch cover failure. The side tanks -- the Derbyshire was a double hull ship -- in way of No 1 imploded, so we know they were intact when the ship sank. Most importantly, the No 1 hatch covers were pushed in, not out. If No 1 hold had flooded before the hatch covers failed, the hatch covers would have been lifted off the coaming by the increase in internal pressure. This has actually occurred in a number of bulk carrier casualties. Faulkner is almost certainly correct. The primary cause of this sinking was that the hatch covers were too flimsy for the low freeboard.

This report essentially exonerating the design of the ship and blaming the crew for the casualty was met with so much outrage that the UK government was forced to re-open the case resulting in yet another report in April of 2000. See RFI entry below.

The Assessors report includes some good pictures of the ship under construction and some sketches of the structural detailing, (which for the most part is much better than current practice). But the report is sadly deficient in hard ship data; hull form, compartmentation, scantlings, which would allow others to replicate the Assessors's calculations, or do their own analyses. The Assessors report does includes a capacity plan but it is unreadable and a very rough sketch of the midship section.

A start but we need much more. Every ship casualty investigation should at a minimum contain, hull offsets, a general arrangement plan, a real midship section drawing, shell expansion, construction profile, forebody, aftbody construction plans, lightweight distribution, a complete loading pattern including bunkers, a mooring plan, engine room arrangement, and hatch cover structure drawing (bulk carriers), preferably in a vector based, machine readable form.


source LINK
type L
volume
material
dead
link http://www.youtube.com/watch?v=fAhaP53wkM0

Simulation which follows the Assessors' report that there was a gradual flooding of the forepeak spaces before the No 1 hatch cover failed. This is controversial to say the least. The simulation is clearly inaccurate in two areas:

  1. The loss of the Bosuns Store hatch cover is assumed to occur very early in the process. This may have happened, but the cover did not fly off by itself as the simulation indicates. It was wrenched off by something hitting its aft coaming, possibly part of the starboard windless.
  2. The explosion of the Forward fuel oil tank and the cofferdam (not depicted) is not shown. This implies that these two spaces flooded very rapidly not in the gradual manner that the simulation portrays. The Forepeak tank may have filled slowly although this is far from certain; but these other two spaces certainly did not. CTX suspects that these spaces flooded as the ship was already sinking.

The simulation implies without saying so that the flooding of the Bosuns Store was a critically important step in the process. In fact, even if this did happen as the simulation shows, this would have decreased the forward freeboard which initially was at least 0.75m above the legal minimum by about 0.4m.

The voice over clearly implies that the primary cause of the sinking was the flooding of the forepeak spaces. In fact, the crucially important event was the failure of the forward cargo hatch covers which Faulkner shows would likely have been badly over-stressed without any flooding at all.

The portion of the simulation following the No 1 hold flooding is accepted by most analysts as the most likely scenario.

Unfortunately, this simulation, commissioned by the Assessors, has been widely circulated; and in most people's mind is the definitive description of the loss of the Derbyshire.


source LINK
type L
volume
material
dead
link http://users.skynet.be/p.woinin/sderby.htm

A stinging critique of the Assessors RFI report by a very experienced and objective bulk carrier captain. Must reading.


source LINK
type L
volume
material
dead
link http://www.nationalarchives.gov.uk/ERORecords/MT/205/3/index.htm

This site contains the evidence and testimony received by Justice Coleman in the Re-opened Formal Investigation (RFI) hearings. The RFI commissioned a number of studies, the two most important of which were:

  1. A hindsight of Typhoon Orchid by a Dr. Cardone which resulted in Cardone's estimate of the maximum wave spectrum that the Derbyshire faced in Orchid. Cardone gave his estimate a +/-11% error margin.
  2. A series of 1:65 scale model tests in the Marin wave tank using Dr. Cardone's wave spectrum.
In combination this research indicated that the probability of the No 1 hatch cover being over-stressed was critically dependent on the freeboard forward, and that this probability was nil to 4% (with +11% error margin) with no flooding, nil to 31% (with +11% error margin) with the Bosun Store flooded and 73% to 99% (with the +11% error margin) with both the Bosun Store and Forepeak tank flooded. Faulkner argued that the linear theory used by Cardone seriously under-predicted extreme events --- often called rogue waves, or freak waves in the popular press --- and it was precisely these extreme events that we were interested in. He also pointed out that such that waves were more likely in a tropical storm that was following a looping, erratic course. He supported his arguments with examples from the North Sea, Gulf of Mexico, and the Sea of Japan. His position was largely dismissed on the grounds that he was a naval architect not an oceanographer. Since 2000 we have accumulated a lot more data on ocean wave heights and slopes, mostly from satellites, and found indeed that the frequency of these extreme event is much higher than predicted by the standard ocean wave models.

The Marin tests also indicated that the Bosuns Store and the Forepeak tank would have flooded faster through the damaged vents then standard hydraulic equations predicted. It is not clear how accurately these 1:65 scale tests replicated full scale conditions in this regard.

The RFI site includes submittals by the Derbyshire Family Association (DFA), Lloyds Register (LR), and the UK Department of Transport, Environment and Regions (DETR).

The DFA submittal is largely aimed at exonerating the crew. The DFA accepts the Cardone/Marin argument that there probably was some flooding forward prior to the hatch cover collapse. But argues that the hatch covers were clearly under-designed. The DFA points out that even using linear wave theory and a 12.5m significant wave height (severe but not all that unusual), if the ship had been loaded down to her legal loadline according to the Marin numbers she would have had a 4% to 40% probability of the forward hatch covers failing. The DFA calls for all kinds of improvements in bulk carrier design, most importantly, a raised forecastle, and much stronger hatch covers.

The LR submittal is revealing. Lloyds Register claims in strongest terms there was nothing wrong with the ship or its design. The hatch cover design was more than adequate. There would have been no problem without flooding. Picking and choosing from the Cardone/Marin numbers. according to LR, this was a "zero probability event". LR statistics show that the design of the vents and other deck equipment was robust and solid. LR agrees with the Assessors as to the cause of the flooding (read crew). But adds a strong criticism of Captain Underhill for following the Ocean Routes recommendations. It urges Justice Coleman to publicly castigate Professor Faulkner for questioning IMO and Class Rules. LR makes the following very strong statement.

The Derbyshire was, therefore, desperately unlucky. Taking all the above factors into account, it seems highly improbable that such a combination of worst cases will ever occur again and it may be concluded that the loss of the Derbyshire was a unique occurrence.
What makes this statement so astonishing is that it was made in July, 2000, after we had experienced dozens of bulk carrier losses in circumstances similiar to the Derbyshire. The CTX database has 44 loaded, large bulk carrier sinkings through 1989.
Loaded Bulk Carrier Sinkings over 50,000 dwt, 1969-1989
Ship Name Date Dead Cgo
bolivar maru 19690105 31 OI
hae dang wha 19800723 29 OI
theomitor 19800813 0 OI
derbyshire 19800909 44 OI
onomichi maru 19801230 0 CL
deifovos 19810125 9 OI
academy star 19820319 0 CL
hope star 19850122 0 OI
arctic career 19850623 27 OI
alexandros f 19860507 0 OI
testarossa 19870113 30 OI
cathay seatrade 19870113 27 OI
orient pioneer 19900107 0 OI
alexandre p 19900315 24 OI
azalea 19900322 4 OI
silimna 19900512 0 OI
tao yuan hai 19900523 0 OI
petingo 19900709 6 OI
pasithea 19900804 31 OI
algarrobo 19900919 32 OI
gallant dragon 19901023 0 OI
protektor 19910111 33 OI
continental lotus 19910121 38 OI
salvia 19910209 0 OI
vasso 19910404 0 OI
starfish 19910408 0 OI
mineral diamond 19910417 26 OI
berlisa 19910520 0 OI
manila transporter 19910707 26 OI
melete 19910824 25 OI
sonata 19911113 0 OI
entrust faith 19911127 0 OI
karadeniz s 19920330 0 OI
great eagle 19920524 0 OI
daeyang honey 19921022 28 OI
nagos 19930526 17 CL
protoklitos 4 19930618 0 OI
marika 19940101 36 OI
kamari 19940225 0 OI
apollo sea 19940620 36 OI
iron antonis 19940903 24 OI
seafaith 19960217 19 OI
iolcos victory 19960914 5 OI
peace 19990128 0 SS
These 44 casualties killed 607 crew. In some of these casualties, the circumstances were eerily similar to the Derbyshire. See the Pasithea and Marika for examples.

The CTX Casualty database contains 12 bulk carrier casualties in which we know the hatch covers were implicated, usually from the Master's distress signal.

DATE SHIP DEAD
19780105 Chandragupta 69
19800909 Derbyshire 44
19801127 Sandalion 0
19810307 Mezada 24
19812297 Marina di Equa 30
19830212 Marine Electric 31
19870429 Skipper I 0
19930526 Nagos 17
19940203 Christinaki 27
19960119 Bluenorth 0
19990214 Bright Century ?
20011222 Christopher 27
These 12 casualties killed 269 crew. We can be confident that weak hatch covers had in hand in a sizable portion of the other 1821 bulk carrier/OBO fatalities in the database.

The LR submittal is an example of the standard reaction by a Classification Society to just about any casualty, although I must say it is even more callous than most. It is must reading for anyone who wants to understand what the Classification Societies really care about.

The DETR submission is a strange document. The first half is devoted to defending the design and construction of the Derbyshire including the ICLL 1966's decision to lower freeboard requirements. The second half calls for all kinds of improvements in the Class Rules and is very critical of IACS and its members for taking a lax view of safety.

The site includes a document called Answers to AG's Draft Questions which compares the answers of these three parties (plus Bibby Lines) to a set of obvious questions. It makes interesting reading.


source CTX
type D
volume 2000T
material B
dead 44
link

The Derbyshire was a double hull OBO; but she had not traded in oil since October, 1979, nearly a year before she foundered. There is no evidence that her carriage of oil had anything to do with her sinking. She had done her 1st Special Survey docking in April, 1980 at Sasebo, where some repairs in way of the pump room bulkhead, Frame 65, were made.

The ship, loaded with iron ore, sank in Typhoon Orchid, about 350 miles SE of Japan. The course she followed had been recommended by Ocean Routes, a commercial ship routing service hired by the ship's charterers. Orchid was not a particularly strong hurricane in terms of wind. It appears the central pressure went no lower than 962 mb with max winds of around 85 knots. However, the storm did follow a long, looping track which could be expected to generate very large, confused waves.

The Derbyshire was UK flag, with a mostly Liverpool based crew, built in the UK classed by Lloyds Register (LR). Initially the UK government refused to investigate, saying "overwhelmed by weather" and nothing could be learned. The UK government had a substantial stake in this case. The yard, Swan Hunter, which built the ship had been privatized and, in order to make the deal work, the UK government idemnified the new yard owners against any claims arising from the ships that the yard had built.

However, Peter Ridyard, a father of one of the crew, was an experienced ship surveyor. He didn't buy the story that a nearly new ship (she was four years old, one year of which she had been laid up) should be "overwhelmed" by a not particularly strong typhoon. Ridyard, being a member of the fraternity, knew that this class of ship had a major structural defect. The problem was that the longitudinal structure was discontinuous at the pump room bulkhead, Frame 65. This is an extremely dangerous practice which is still allowed by Class Rules. He knew these ships had a bad cracking problem in this area, and figured this was the real reason that the Derbyshire sank. In March 1982, a sister ship, the Tyne Bridge, experienced severe cracking at frame 65 just forward of the accommodations. The ship was in ballast the North Sea. The Captain issued a mayday but was able to get to a shelter.

Still Ridyard's efforts went nowhere. In March 1986, the UK DOT prodded by increasing publicity issued a report saying again there was not enough evidence to make any determination, and dropped it. Their timing was very bad. Eight months later another sistership, the Kowloon Bridge, developed severe cracking at Frame 65. She then managed to lose her rudder and ended up aground on the south coast of Ireland. She broke her back, the break occuring near Frame 65. It was then learned that the ship has had massive girders welded over the deck near Frame 65, a fact that nobody (yard, Class or owner) had seen fit to make public until then.

Between October of 1987 and March 1988, the UK government held yet another investigation, but came up with the same answer:

For the reasons stated in this report, the Court finds that the Derbyshire was probably overwhelmed by the forces of nature in Typhoon Orchid, possibly after getting beam onto the wind and sea off Okinawa in darkness of the night 9th/10th September 1980 with the loss of the 44 lives. The evidence available does not support any firmer conclusions.

The relatives of the crew organized as the Derbyshire Families Association (DFA) were outraged and went public big time. They were able to enlist the support of Tyne Tees Television and the ITF which paid for an underwater survey in 1994. The UK government had argued that a search for the Derbyshire would be fruitless. It was not known exactly where she had sunk; the water depth was over 4000 meters; and technology to obtain clear photographic evidence did not exist. The ITF effort had the good luck to find the ship only 23 hours after starting the search and took some very interesting pictures; but deteriorating weather and limited money curtailed the picture taking.

The UK now convened yet another investigation under Lord Donaldson. Donaldson concluded that another underwater survey was required. On this survey, finished in April 1997, close to 136,000 photographs were taken. The quality of some of the pictures was amazingly good. To just about everybody's surprise, the pictures did not support the cracking at Frame 65 theory. Despite the fact that the ship had needed to make some repairs to the Frame 65 structure in April, 1980, the hull forward and aft of Frame 65 were fairly close together on the bottom, which would have been extremely unlikely if the ship had split at Frame 65 on or near the surface. Even more tellingly, the port and starboard slop tanks just aft of Frame 65 had imploded. They must have been intact when the ship started her dive. This could not have happened if there had been cracking at Frame 65 prior to her sinking. The cause had to be elsewhere.

The UK government took a stab at that good old stand-by, blame-the-crew. The DOT concocted a scenario of flooding of the bow Bosuns Store due to the crew's leaving the hatch open. This was based on a photo. which showed the forward Bosuns Store hatch cover missing and a line trailing from the store out of frame. The government duly issued a report in March, 1998 known as the UK/EC Assessor's Report (see Entry above) implying crew negligence was the cause of the loss.

The idea that three days before reaching port, facing a major typhoon, the crew would pull a line partially on deck and then just leave it there with the hatch open was willful idiocy. Crews don't pull mooring lines out 3 days before reaching port even if they know the weather would be perfectly calm for the entire period. When they pull a line out, they don't leave it half in and half out. And even the most suicidal crew would not leave the hatch cover open going into a typhoon.

A Chief Mate who had served on the Derbyshire quickly stepped forward, and pointed out that the practice on the ship (and many ships) when storing the lines, was they would be lashed together and the end of the last length of line fastened to the underside of the hatch cover. This made it easy to pull out all the lines when needed. However, the hatch cover failed, the line would pay out as the hatch cover separated from the hatch. The same picture. showed the hatch had been blasted at the aft end destroying the aft hatch coaming. This was clearly the most likely cause of the hatch cover failing.

In any event, flooding of the 700 m3 Bosuns Store by itself would not have sunk the ship. It would only have reduced the forward freeboard by 0.39 m.

In the Faulkner reference listed above, Faulkner argues that the loss was due to the collapse of the forwardmost hatch cover. Under Class Rules, the hatch covers are designed to take a pressure of only 1.75 meters of water on top of them. Faulkner found that the Derbyshire's hatch covers met this rule. He calculated they would fail at a head of about 4 meters. However, Faulkner's analysis also indicated that, in the conditions that the ship encountered, there was a high probability of the ship's hitting a wave large enough to collapse the ship's forward hatch covers, and quickly flood the forwardmost hold. After that the ship was doomed. He recommended that the strength of the hatch covers be increased by a factor of 3, which would not have required very much steel.

It is quite possible that bow flooding could have contributed to the collapse of the hatch cover. There could have been nil, and the ship encountered a large enough wave. There could have been a lot, in which case the wave required to collapse the hatch cover could have been a bit smaller. Either way the hatch cover was not strong enough.

Faulkner's analysis also showed that the ship could easily have encountered a wave train that would have exceeded her IACS required bending moment by 80%. In theory, Class Rules are based on designing to a wave that the ship will encounter on average once in every twenty years. But Faulkner and others have shown that the way Class determines this wave is flawed, and even this very questionable criteria is not met.

One of the more minor problems with our Class Rules designed ships is that they have so many things wrong with them, that it is often impossible to figure out which is most responsible in a given casualty. The Derbyshire families had good reason to suspect cracking at Frame 65. The hatch covers are far too weak. The hull longitudinal strength is at best marginal. The lack of a raised forecastle is indefensible.

Faulkner's results meant not just the Derbyshire was doomed, but other vessels as well. In fact, between January 1990 and February 1998, eight very large ships loaded with iron ore simply disappeared.

One of those ships was the Pasithea. There was a video on youtube taken from the bridge of the Pasithea. She is loaded. Like the Derbyshire and almost all post-70 built bulk carriers, she has no raised forecastle. The weather is not bad, but a big swell is running. Crew are walking around on deck, but only on the aft portion. The ship is regularly submerging her bow and water is washing over the deck as far aft as the kingposts. Sadly, the video is no longer available. The Pasithea sank off Japan almost exactly ten years after the Derbyshire and under very similar cirumstances. The entire crew, 25 Greeks and 7 Filipinos, were killed. But this time there was no real investigation. Despite the fact that the Derbyshire showed that we have the technology to not only find wrecks in the deep ocean, but also take insightful pictures, CTX knows of only three attempts to underwater survey a wreck since the Derbyshire, and none nearly as extensive.

Faulkner's arguments did not find favor with officialdom. The UK government was torn between placating the Derbyshire Families and their public support, and defending Swan Hunter. Faulkner was forced off the UK/EC investigation prior to the publication of the Assessors' Report. Glenda Jackson, a minister at DOT (which by then had been given the much more trendy name of Department of Environment, Transport and the Regions) complained

Professor Faulkner's active promotion of the theory that the Derbyshire sank when her hatch covers gave way is a matter of dismay for the Department. Officials have written to Professor Faulkner to convey the Department's displeasure in his untimely interventions. In response he has offered his resignation as Assessor.
If this is the response of the UK regulators, what can we expect from flag states such as Panama?

The Assessors' report generated yet another outburst of public outrage and the UK government was forced to re-open the case. This resulted in the publication of yet another report known at the RFI (Re-opened Formal Investigation) report in April of 2000. CTX has not seen this report but apparently it exonerated the crew, accepted the Cardone/Marin argument that the flooding forward was crucial, and made some weak recommendations for strengthening bulk carriers.

CTX has not had a chance to examine the Cardone/Marin analyses that indicate that a 1.5m change in draft forward changes everything as far as the hatch covers are concerned, flipping the probability of hatch cover failure from near zero to over 75%; but this simply doesn't make sense. Freeboard is important; but not that important. And if it is true, it is a compelling argument for more freeboard which apparently Justice Coleman inconsistently ignored.

IMO responded to the Derbyshire by passing some badly flawed rules for improving forepeak pumping arrangements which will do nothing when a cargo hatch cover fails. IACS has responded to the Derbyshire with a modest strengthening of the hatch cover rules, far less than Faulkner's analysis calls for. Neither required more freeboard nor a raised forecastle. The suggestion to either retrofit stronger hatch covers or increase the minimum freeboard of existing bulk carriers has been ignored. There has been no real change in the longitudinal strength rules. See the Christopher sinking for an example of the results of this relaxed attitude.

Bulk carrier losses are distressingly common. The only thing unusual about the Derbyshire was (a) the crew's relatives were able to get this casualty into the public eye and actually investigated, and (b) one of the investigators proved to be too honest to toe the official line. If the industry's normal regulatory apparatus had had its way, as it certainly would have if the crew had been say Greek/Filipino, the whole thing would have stopped with the first finding of "over-whelmed by weather".

This casualty produced a bunker oil spill. The Japanese SAR aircraft found oil bubbling to the surface near the Derbyshire's last known position. It was this information that allowed the first under-water survey to find the wreck so quickly. CTX as yet has no info on the bunkers on-board. The guess at 2000 tons is based on a normal reserve plus enough fuel to get to Singapore which is the closest location of cheap bunkers. We need the actual figure.