Back to Casualty List | Search The Casualty Database
Precis File
SHIP NAME: Rocknes KEY: NUM. ENTRIES: 6
source GISIS
type A
volume
material
dead 18
link

Summary of events On the weekend of 17 and 18 January 2004, the bulk carrier ROCKNES loaded a cargo of gravel in the South Norwegian port of Eikefet, north of Bergen. After the loading works had been completed on Sunday evening, the vessel left Eikefet shortly before 09:00h on Monday 19 January 2004, with a pilot on board and proceeded to bunker in Skalevik near Bergen, where she moored at 11:40h. Four hours later, at about 15:40h, mv ROCKNES continued its voyage, still with a pilot on board, heading for the intended port of discharge in Emden, Germany. At about 16:23h, mv ROCKNES changed course to port in order to enter Vatlestraumen. At about 16:27h the vessel touched ground with her starboard side in the narrow passage; immediately she developed a list to starboard that could not be stopped. Within a few minutes mv ROCKNES capsized completely. Twelve of the thirty persons on board could be recovered alive as a result of the rescue measures initiated immediately. Three of them were freed from the hull of the vessel floating keel upwards after several hours. Escaping fuel polluted Vatlestraumen and the adjacent coastline. The Norwegian Coastal Authority collected a total of 1.291 tons of emulsified water which has been determined to be equivalent to 227,5 tons of IFO 380. After the rescue attempts were discontinued, the vessel was towed into a bay on Agotnes. The hull of the vessel was examined and it was decided to upright mv ROCKNES again. From 17 March 2004 onwards work proceeded on turning the vessel; on 29 March 2004 it was lying upright again. On 5 April 2004 mv ROCKNES was towed to the Bergen Mekaniske yard at Laksevag near Bergen. The vessel was inspected and a decision was taken to reinstate it. To this end mv ROCKNES was towed to Poland for repairs and modifications. It was initially estimated that the necessary repair works would take more than one and a half years.

The fully loaded 17765gt bulk carrier ROCKNESS had left Bergen, Norway en route for Emden, Germany. Shortly after altering course to transit the Vatlestraumen passage, the bulk carrier struck bottom on her starboard side. The vessel immediately developed a list to starboard which could not be stopped, and within a few minutes had capsized. Only 12 of the 30 crew onboard at the time were rescued. Approximately 227 tones of IFO 380 fuel escaped from the wreck and polluted the area and adjacent coastline.

Originally built as a self loading bulk carrier, the vessel had an 80m swivel type boom fitted forward amidships to discharge its cargo, which created a blind arc forward. Subsequent modification added a flexible fall pipe system to enable the vessel to deposit its cargo selectively on the sea bottom. The associated structure and modifications increased the blind arc forward, and raised the vessel’s centre of gravity (CofG). On the day of the accident, the vessel had too high a CofG, and the cargo had not been trimmed. Had the vessel been loaded correctly, it might still have capsized, but more slowly thereby allowing the crew more time to escape.

The pilot did not have the latest chart available, and therefore was unaware of the extent of the hazard area to starboard of the planned track. Further, he was navigating by eye which, given the blind arcs forward, was ineffective. A combination of visual and radar navigation would have been more effective.

The stability of ROCKNESS was independently assessed by the Norwegian Maritime Directorate and resulted in Norway making proposals to IMO’s 48th SLF Sub-Committee on sub-division, calculation of attained stability, and use of the cargo computer.

The navigation Authority subsequently adjusted the limits of the channel's sector light to take account of the revised seabed topography, and laid a buoy to mark the extremity of the hazard.

When confronted with the vessel’s blind arcs forward, the pilot did not modify his plan to conduct visual navigation by involving the rest of the bridge to assist with track monitoring both visually and by radar. Neither, had the master taken steps to plan a suitable passage or to ensure that he and his staff were adequately monitoring the safety of the ship.


source Lusk, Laying the Blame
type A
volume
material
dead 18
link

State Partly Liable but Claims also Found to be Negligent Captain Barry M. Lusk, Lusk Hydrographic Expertise (Canada)

Although Norwegian authorities immediately investigated the disaster, responsibility for the accident remained undetermined, leading to the reopening of the case in February 2009. In the April edition of Hydro international we discussed this in ‘The 2004 Rocknes Drama, Past and Present'. In May the court determined the verdict.

As well as the death of 18 members of the crew, the incident also led to 450 tons of fuel oil being spilt along a 15-km stretch of the Norwegian coastline. The claimants argued that, had the shoal been properly reported, the pilot would have been aware of its presence and the accident avoided. The opinions presented here may allow the reader to arrive at some conclusions that may not agree completely with those determined by the court: many negative factors combined to create an accident waiting to happen.

The owner of the M/V Rocknes knew that she was inherently unstable partially due to the 3,000 tons of new ironwork on her forward deck and the shape of her cargo holds (see Figure 1). Ballast tanks that had been retrofitted to counter this instability were slow to fill and difficult to maintain. Forward visibility was compromised, even though Lloyds Register of Shipping [actually Germanisher Lloyd] had approved the refitted alterations. Further, the ship was difficult to steer, perhaps because she lacked a keel and had a tendency to heel over substantially and slide sideways when any amount of rudder was applied. To compound these problems, she was incorrectly loaded and her cargo holds were not trimmed prior to setting to sea.

Although the pilot knew Vatlestraumen Narrows very well and was capable of navigating this passage without the aid of radar, bearing and distances or measured sight, the ship was out of mid-channel and too far to starboard (see Figure 2). As the state argued, the charts did not contain any information to indicate that, at the location of the grounding, depth was more than 10 metres.

During the reconfiguration from bulk carrier to flexible fall pipe ship, steel structures were added to the deck forward of the bridge and over much of the seven cargo holds. It is reported that 3,000 tons of weight had been added. To compensate for the added weight, all of which was above the original centre of gravity, ballast tanks were added to each side of the ship. [not true, sponson tanks were only added after the casualty]. In addition, the six holds used for the storage of its cargo were altered in such a way that a conveyer-belt system could remove the cargo from each hold and deliver it to the flexible pipe arrangement that was situated over cargo hold number five. Each of the cargo holds was triangularly shaped with the apex of the triangle on the centreline of the ship, pointing downwards. This placed the majority of the cargo in the holds well above the hull waterline, decreasing the metacentric height (distance between the ship's metacentre and centre of gravity, referred to as GM) unless ballasting and fuel compensations were effected correctly.

Further to the stability problem was a concern that visibility forward from the bridge was not acceptable. In order to see forward at all, the navigation officer had to walk to the port or starboard side of the bridge. Although the ship had passed inspection regarding forward visibility, the fact remains that forward visibility from bridge midships was almost non-existent. Various navigation officers had expressed concern that visibility was only unobstructed from each side of the bridge.

Regarding the cargo on this ill-fated voyage, the loading conveyor at the Norwegian port of Eikefet was stationary and was of insufficient length to load the far side of the ship (the starboard side). In this case, the port side (against the dock) was loaded and the starboard side after settling was not. Because the shore conveyor was stationary, the ship had to be reefed forward and aft filling all six holds individually. The important facts here are: (1) the ship left the dock without having its cargo holds trimmed, and as a result less cargo was on the starboard side than the port side; and (2) because of the nature of the loose cargo, it had a tendency to shift or avalanche quickly.

Even as the ship travelled from Eikefet, the captain and pilot (who had boarded the ship for the inside passage run) discussed the steering and stability characteristics, which they had noted to be poor. Difficulties were experienced when normal changes of course were made to the rudder angle. The ship would heel over and slide in the direction the ship had been travelling prior to the alteration of course. The captain expressed some concern that the ship's centre of gravity was offset and therefore its general stability was not good. He and others found that the only way to minimise this dilemma was to make many small alterations in order to set the ship on the desired course.

The shoal on which the M/V Rocknes grounded in January 2004 was discovered by the Norwegian Chart Authority in 1995, but had not been reported in Norwegian "Notices to Mariners". However, the pilot aboard the ship was traversing Vatlestraumen Narrows by line of sight and personal knowledge. No navigational aids (except those fixed aids ashore) were being used; indeed, it would be unreasonable to expect anything else under these circumstances. The pilot had taken ships such as this through Vatlestraumen Narrows many times, and knew the area well. The shoal on which the ship ran aground was a slight extension to an already-known shallow area that was affixed to the shore low water and correctly marked by a fixed aid ashore. The judge in the case has found that the Hydrographic Office did not notify the maritime community of this 10-m contour extension in an appropriate or timely manner and because the Government appeared not to have followed acceptable procedures in this regard found the Government partially responsible. In my opinion, they were not in any way accountable but Governments have a responsibility to appear to do what is right and they have been ordered to pay some minor costs.

Lack of knowledge of the extension of the 10-m contour was only of consequence because the ship was too far to the starboard side of the channel, due to three different factors.

1. The visibility from the bridge was poor and it was not possible to cover both the starboard and port sides of the ship, necessary to see whether the Rocknes was mid-channel as she should have been.

2. The Rocknes was attempting to remain well astern of a ship ahead. This affected the attention of the pilot and distracted him from other navigational duties, especially as he had to walk from one side of the bridge to the other to see this ship and see where the Rocknes was going.

3. The steering characteristics of the Rocknes were so temperamental that it was difficult to maintain or adjust to a chosen course.

As the ship was too far to the starboard side of the channel, she touched the edge of the 10-m contour line and punctured three areas of the hull on the starboard side. According to reports, the forward two were substantially breached and the third was partially breached. From photographs of the hull ruptures, none of the damage appears excessive to me. However, water in the two forward ruptured holds compromised an already unstable ship. The improperly loaded cargo was shifted to the starboard side in an avalanche fashion, capsizing the ship within four minutes (estimated by an observer) of the grounding.

In May 2009, the judge in the Norwegian court at the trial of the ship owners and their insurance agents against the Government of Norway and its Hydrographic Service found that the State was liable in not reporting the shoal, as I have already said. However, negligence on the part of the owners and the operators of the ship (the ship's captain and pilot, primarily) meant that the claimants were held responsible for 96% of the cost of the resultant damage and ultimately the grounding itself. These claims were for the damage to the ship and the resultant repairs, the cost of the oil clean-up on the shoreline and the various claims made by the families of those who died in the disaster. Although the ship ran aground in the red sector of an extension to the natural shoaling close to the shore, the existence of which had not been properly promulgated to the maritime community, the existence or non-existence of the shoaling was not the primary factor in the grounding itself. Assessment of financial responsibility by the court acknowledges this fact.

Further Reading Weber Smit, A., 2009: The 2004 Rocknes Drama. Past and Present. Hydro international, 13 (3), 30-31.


source LINK
type D
volume
material
dead 18
link http://www.theartofdredging.com/Investigation_Report_18-04.pdf

This is the official flag state (Antigua and Barbuda) report. But it was a joint report with Germany (state with substantial interest), and appears to have been written mainly by the Germans. It is packed with facts about the voyage, and is must reading for anyone interested in this casualty. But it almost willfully ignores the ship's grievous design faults and focuses on the bridge management, in discussing why the pilot ended up further starboard than he intended. It says little about the rapidity of the capsize or its causes. There is no analysis of either the intact or damaged stability despite this being a crucial factor in the casualty.

The Antigua/Germany report can also be found at epub.sub.uni-hamburg.de/epub/volltexte/2008/1798/.


source LINK
type D
volume
material
dead
link http://www.theartofdredging.com/Annex_5.pdf

This is Damage Card and Intact Stability calculation submitted to IMO presumably by the flag state Antigua. The Annex 5 document is horribly designed. The two most important damage location numbers are the high point and the low point. Yet these numbers are so buried in an unnecessarily complex arrangement that they rarely end up being submitted. They were not submitted in this case.

Worse the intact stability calc assumes the port and stbd righting arms are symmetric. This is true only if both the ship and the cargo load is symmetric, which is often not the case in capsizes. It certainly was not the case in this casualty.

The estimated metacentric height in this Annex 5 is 0.385 m which is nearly the same as that submitted by the operator (Jebsens) (0.398) to the Norwegian inquest. CTX has not seen the owners calcs but once again symmetry seems to have been assumed. The legally required GM is 0.62 m.


source LINK
type D
volume
material
dead
link http://www.theartofdredging.com/The_capsize_of_MV_Ro_27082a.pdf

This was an independent report done by someone at Bergen University College. When the report was published, Jebsens legally pressured the college to withdraw it. The College then granted the unnamed author a paid leave to publish the report privately, presumably insulating the College from any Jebsen suits. The report is absolute must reading for any one interested in maritime safety.

The Bergen report argues that the stability assumptions used in both the design and operation of the Rocknes were unrealistically [CTX would say fictitously] optimistic. When the Rocknes was converted from a self-unloading bulk carrier to a fancy rock dumping ship, the VCG of the [light?] ship was raise from 9.66 m to 10.91 m, a moonpool and associated equipment on the starboard reduced starboard buoyancy while at the same time shifting the center of gravity to starboard. As a result, the ship had at best very marginal stability.

To meet the statuatory requirements, in part a GM of 0.62m, Germanisher Lloyd (GL), the Classification society, assumed an optimistically high cargo bulk density, the cargo being perfectly trimmed, and perfect ballasting.

Jebsens own submittal which also used a higher than lab bulk density came up with a GM of 0.39, clearly illegal. Yet no one suggested this was an unusual load. The Germanisher Lloyds (GL) design calculations were clearly fictitious, and it is hard to believe GL did not realize this.

It is undisputed that in fact the cargo was not only not trimmed but because the loading conveyor was to short, so the cargo ended up being loaded off-center to port. Not only did this mean that the cargo CG was higher than assumed by the Loadicator, but also the normal ballast required to put the ship at zero heel was not needed. If the ship had ballasted per the book, she would have listed to port.

Putting all this together, the author of the Bergen report estimates the intact GM was between 0.17 m and 0.02 m. More importantly, his calculations indicate that the righting arm at zero degrees was negative, resulting in instability between +/- 2.6 degrees. This is supported not only by the pilot (the only surviving bridge team member) but also by witnesses at the loading terminal, the bunkering terminal and nearly ships who claimed the ship was listing back and forth. Jebsens claimed the witnesses had to be mistaken, since by its calculations this could not have happened. The last part of this claim is certainly true.

It was also common knowledge among those familiar with the ship that any turn has to be made slowly and gradually, lest the turning moments overwhelm the minimal righting moments, resulting in a high list, and an asymmetic waterline which resulted in the ship sliding sidewards. Thus the pilot was faced with traversing a narrow passage with a ship that had extremely limited maneuverability. To make a fairly abrupt 55 degree turn to port (he had to go from 178 to 123), he used only 5 degree rudder. Even then the ship listed to starboard.

To make matters even worse, the combination of the big self-unloading boom and the massive structure above the moonpool, resulted in nil visibility from the wheelhouse. To get around the visibility regulations, GL called the moonpool structure a "lattice" and acted as if it were not there. In fact the only way the pilot could see anything is to move from bridge wing to bridge wing, always being blind on at least one side. The visibility regulations were clearly contravened.

With nil maneuverability and very limited visibility it it hardly surprising that the pilot drifted to starboard of his intended course.

Damage was confined to the turn of the bilge in two wing ballast tanks, and a very small hole (possibly not part of the initial damage) in another tank. The Norwegian inquest, using mainly data supplied by Jebsens, found that, if the ship had had legal stability, she still would have capsized, but more slowly which would probably have cut the death toll a lot. The Bergen report argues that she might have avoided the capsize. It estimates that the equilibrium angle of heel would have been 30 degrees.

But this assumes the cargo did not shift. In fact, eyewitness reports indicate that the heel rate increased dramatically from about 3 deg/min to very roughly 110 degrees/min when the heel reached about 15 degrees. The only way this could happen is the cargo shifted. [More evidence of optimistic design assumptions. To meet the damage stability rules, GL had assumed the cargo would not shift until 45 degree heel.]

In order for the ship to survive, listing at 30 degrees, not only would the ship have had to have been loaded legally from a stability point of view, but the cargo would have had to have been well enough trimmed so the cargo would not have shifted at 30 degrees.

The abrupt change in heel rate was undoubtedly a major factor in the death toll. 3 deg/min for the first five minutes would not have been too alarming. The crew probably felt that the ship was not going to capsize, or at least they had plenty of time.

Preposterously, Jebsens argued that there is no evidence that the ship would have capsized more slowly, if she had been legally loaded. Legalities trumping physics.


source CTX
type D
volume
material
dead 18
link

This casualty has a little of everything:

a) a horribly unsafe design, duly approved by Class in part by using impossibly optimistic assumptions.

b) illegal, unsafe but standard operating procedures which inter alia violated the design assumptions,

c) a determined attempt by the ship's operator to push blame away from itself and put it on the pilot, the crew, and the Norwegian cartographers.

(a), (b) and (c) are commonplace in marine casualties. What is unusual in the case of the Rocknes is we have so much information.

It is also interesting that despite the clear evidence that the ship was both unsafe and illegal, and that the combination of the illegal ship and the clearly illegal operating practices nearly guaranteed a casualty, no criminal charges have been brought against anybody despite 18 people being killed. As far as CTX knows, Germanisher Lloyds has totally escaped any kind of penalty.

The Rocknes was raised and rebuilt, becoming the Nordnes. Despite owner's and Class's claim that the Rocknes was a safe ship, the Nordnes was fitted with additional bulkheads in the ballast tanks, a skeg to improve maneuverability, wider bridge wings, and most tellingly 750 mm wide sponson tanks on either side of the hull. If Germanisher Lloyd had adopted a realistic design philosophy, these sponson tanks would have been required when the ship was originally converted in 2002. But the design philosophy of assuming perfection from equipment and operators permeates all the Class rules.