Summary

Early on the morning of 29 December 1993, the Australian offshore supply vessel Shelf Supporter was discharging a deck cargo of empty skips and a container to North Rankin A platform, 70 miles north-west of Dampier. The weather was fine with the wind from the south-west at 20 knots.

The Master was manoeuvring the vessel, maintaining position, from the after end of the bridge, looking out over the cargo deck.

Seas were breaking over the stern, but conditions were not such as to cause the cancellation of cargo operations.

While a skip was being hoisted to the platform the two-man deck watch aboard Shelf Supporter decided to re-spool a wire used to secure the cargo, onto the winch. While this wire was being released from a block, a sea broke over the stern. Despite a warning from the Master, one of the two men on deck was crushed between the vessel's own rubbish skip and an empty skip when it was moved by the seawater.

Conclusions

It is considered that:

1. The sea conditions at the time of the incident were not such as to warrant the cancellation of cargo operations.
2. There was no apparent reason why the AB should not have heard the warning made by the Master.
3. The AB was apparently caught unawares by the forward movement of the skip.
4. The decision to re-spool the port tugger wire, to prevent it from being damaged, was in accordance with good seamanship but, coupled with an apparent belief that the position immediately forward of the athwartships skip was a safe one, resulted in the AB being in such a position so as to be crushed between the two skips.
5. The act of leaving the port tugger wire lying flaked along the deck, and not re-spooling it on the winch as soon as the first lift was clear, was not in accordance with good seamanship and the delay in re-spooling the wire resulted in the athwartships skip becoming exposed before action was taken.
6. Athwartships stowage, particularly of empty skips and other forms of bin, is undesirable, due to their greater vulnerability to movement by seas breaking on board.
7. All cargo operations should be fully discussed between the platform/rig manager and the supply vessel master, and the order of operations agreed, before the vessel goes alongside.
8. Although time may be at a premium, crane hooks should not be lowered to the supply vessel until such time as the crew on deck have signalled their readiness.

On 14 January 1994 at 0114, Aero Commander 690 aircraft VH-BSS struck the sea while being radar vectored to intercept the Instrument Landing System approach to runway 34 at Sydney (Kingsford-Smith) Airport, NSW. The last recorded position of the aircraft was about 10 miles to the south-east of the airport. At the time of the accident the aircraft was being operated as a cargo charter flight from Canberra to Sydney in accordance with the Instrument Flight Rules. The body of the pilot who was the sole occupant of the aircraft was never recovered. Although wreckage identified as part of the aircraft was located on the seabed shortly after the accident, salvage action was not initially undertaken. This decision was taken after consideration of the known circumstances of the occurrence and of the costs of salvage versus the potential safety benefit that might be gained from examination of the wreckage.

About 18 months after the accident, the wing and tail sections of the aircraft were recovered from the sea by fishermen. As a result, a detailed examination of that wreckage was carried out to assess the validity of the Bureau's original analysis that the airworthiness of the aircraft was unlikely to have been a factor in this accident. No evidence was found of any defect which may have affected the normal operation of the aircraft.

The aircraft descended below the altitude it had been cleared to by air traffic control. From the evidence available it was determined that the circumstances of this accident were consistent with controlled flight into the sea.

No safety recommendations resulted from this investigation.

Summary

The New Zealand registered roll on/roll off, gas turbine/electric ship Union Rotorua was about 27 miles south of Sydney Heads, en route for Melbourne, when a fire broke out in the gas turbine house.

It was established that the seat of the fire was in one of the cubicles of the 6.6 kilovolt switchboard, which distributed power from the gas turbine generator to the propulsion motors and auxiliary electrical systems.

Initial attempts to fight the fire with extinguishers proved to be ineffective and a decision was made to utilise the fixed firefighting medium and flood the space with CO2.

This was effective in extinguishing the fire. With the high voltage switchboard out of commission, the vessel was towed to Sydney, where it remained for three months undergoing repairs.

Conclusions

1. It was not possible to establish the exact cause of the fire, although evidence indicates that it probably started in the yellow phase of the main generator circuit breaker at the lower "tulip", connecting the generator input to the circuit breaker.
2. Initiation of the fire at the "tulip" connector may have been due to high electrical resistance caused by one or more of the following:
   • Loss of spring pressure (due to age or other causes) behind the connecting "fingers" of the "tulip".
   • Oxidation of the contact surfaces due to loss of silver plating on the contacts.
   • Mechanical fracture, due to vibration, of the end of the circuit breaker's connecting post at the point of change of cross-sectional area.
   • Loosening or fracture (due to vibration) of the socket-head cap screw securing the aluminium alloy collar mounting the connecting "fingers", to the end of the copper post.
3. Although the Electrician's workbook noted, in June 1991, that the contacts on the "Main HT breaker" needed re-silvering, there was no indication that this was ever carried out.
4. The fire spread to the adjoining cubicle containing the circuit breaker for the no.1 ship's service transformer. This may have been due to ionisation of the atmosphere by products of combustion from the fire in the main generator circuit breaker cubicle.
5. The Inspector considers that the delay in shutting down the gas turbine generator, after the fire was discovered, was excessive, but that the actions of the ship's officers and crew in tackling the fire were prompt and correct in all other respects.
6. The emergency response was effective and demonstrated planning and procedures of a high standard.

Summary

On 10 December 1993, the Australian flag tanker Osco Star was loading a cargo of petroleum products at the Shell refinery Geelong (Corio). A little after midnight, as the officer of the off-going watch was checking the deck, he opened the sighting port to no.4 port cargo tank and heard the rush of liquid falling into what should have been an empty tank. On checking he found that the gas oil cargo in no.3 centre tank was flowing into no.4 port.

Cargo operations were suspended and an inspection made of the area. It was subsequently discovered that gas oil had also entered the trunk-way to no.4 double bottom ballast tank.

Initially, it was considered that the ship had been over-stressed during cargo operations. However, by late morning on the 11 th, it was accepted that the tank structure had suffered damage due to over-pressurisation, caused by the failure of the pressure/ vacuum relief valve on no.3 centre cargo tank, a valve which is designed to vent the displaced gas in the tank when loading, or to relieve vacuum when discharging.

Substantial damage was sustained by the vessel's structure, with splits occurring between no.3 centre cargo tank and no.4 port wing tank and the trunk-way to no.4 double-bottom ballast tank. The top of the bulkhead between nos.3 and 4 centre cargo tanks was set back approximately 750mm and the main deck above the tank was set up by 116mm.

The Marine Incident Investigation Unit investigated the incident under the provisions of the Navigation (Marine Casualty) Regulations.

Conclusions

1. The structural damage experienced by the vessel, in the vicinity of no.3 centre and no.4 port wing cargo tanks, was caused by an excessive buildup of gas pressure in no.3 centre tank as the tank was loaded. This was, in turn, caused by the failure of the single PIV valve on no.3 centre tank to open and vent the tank.
2. The PIV valve met the International Maritime Organisation standard and was of proven design. However, the cast iron used in the PIV valves on Osco Star resulted in a build-up of rust and cargo residues in the critical area between the valve housing and the edge of the valve disk and also between the valve housing and the top cone.
3. Due to the fact that the ship made relatively few voyages in ballast, the opportunity to overhaul the P/V valves at sea was limited.
4. The Chief Mate knew of a deficiency in no.3 centre tank PN valve but did not ensure that the valve was repaired or that alternative open loading procedures were followed before loading cargo to that tank. It is probable that the need to ensure that the valve was repaired slipped his mind.
5. Other members of the ship's staff were aware of the defect in no.3 centre PIV valve, but no action was taken to rectify the situation, and only a limited and low-key reminder was given to the officer of the watch, and no alternative open loading procedure was suggested.
6. The Third Mate, when handing over his watch, indicated that the PN valve on no.3 centre tank was functioning correctly, without having previously established that this was the case.
7. There was a break-down in procedures in addressing the repair of no.3 centre PN valve and a failure of communications concerning the importance of ensuring the valve worked correctly.
8. The information obtained during the course of this investigation, with particular reference to the report of the Salvage Association, UK, would demonstrate that incidents of over-pressurisation of cargo and ballast tanks, with the associated danger to life and the risk of pollution, occur very much more frequently than is commonly appreciated throughout the maritime industry.

Summary

In the early hours of 29 October 1993, the offshore supply vessel Pacific Commander manoeuvred stern-to the semi-submersible drilling rig Atwood Falcon to back-load some sections of casing. The Master was advised the cargo operation would be of short duration and, as a result, he chose not to use an anchor, making fast with two lines aft and maintaining position, at right angles to the rig, with the bow thruster.

After the planned cargo operation was completed and some further equipment transferred, Atwood Falcon then decided to take the opportunity to top up with cement from Pacific Commander, thereby extending the cargo operation significantly. Weather conditions were good, and the Master continued to maintain position by use of the bow thruster. In the early afternoon, the Mate was unable to maintain the position and the Master decided to suspend operations and let go. During the operation of letting go, a nylon mooring line, jammed on the bitts, parted under strain, injuring the Mate.

Conclusions

It is considered that:

1. The information provided to the Master prior to Pacific Commander going alongside Atwood Falcon gave no indication that the operation would be other than relatively short.
2. Based on the initial information provided to him and his experience in the industry, the Master's decision not to use an anchor on that particular occasion was not unreasonable. However, an earlier exercising of his power of discretion to suspend operations, while he remoored using an anchor, would have been appropriate.
3. Full length nylon mooring ropes should not be used, because of the inherent dangers of their elasticity.
4. The configuration of the capstans and bitts on Pacific Commander, with capstans at a higher level than the bitts and close in fore and aft alignment, is such that the mooring lines are prone to becoming jammed when turned up on both and under tension.
5. As changes in plans can occur at short notice, and as the operations of anchoring and weighing anchor are routine, provided water depth permits, an anchor should be used during cargo operations as a standard procedure for vessels not equipped with directional/azimuth thrusters, unless the operation is limited to a single lift.
6. The planning of cargo transfer operations by rig staff was deficient and subject to ad hoc alterations without proper consultation with the Master of the support vessel.

Summary

In November 1993, two reports of close quarter incidents were brought to the attention of the Inspector of Marine Accidents. In both cases the navigation of the two ships concerned was under the charge of pilots licensed for the Great Barrier Reef inner route. The four ships involved were "regulated" ships within the meaning of the Great Barrier Reef Marine Park Act 1975. Under the Act any ship over 70m, navigating the Reef between Cape York and Cairns (Low Isles), must take a duly licensed pilot.

The Inspector regarded the reports. as "incidents" in that each one was, on the face of it, an event as a result of which serious damage to a ship or the environment might reasonably have occurred, and it was reasonably suspected that the safety of ships' personnel were imperilled.

Incident 1

On 2 November 1993, the south-bound Philippine flag bulk carrier Blossom Forever was slowly overtaking the Indian flag bulk carrier Pearl Prosperity in an area of the Great Barrier Reef where the maximum width of fairway reduced from about 1.5 miles to about 1 mile. The differential in speed meant that the overtaking manoeuvre would take about 45 minutes or about 10 miles to complete.

Both ships, in ballast en route for Hay Point, were of almost identical size, marginally over 180m in length, with deadweight tonnages of 38,852 tonnes and 34,554 tonnes respectively.

The two pilots had been in contact by VHF and it was mutually agreed that Blossom Forever would overtake on Pearl Prosperity's port side. The distance by radar between the two vessels reduced to 143m, with the overtaking vessel between 30 and 60 degrees on the Pearl Prosperity's port quarter. The Pearl Prosperity's Pilot considered that the vessel was unnecessarily close and the passing distance would be less than a cable (185m). The ship's master commented on the fact to the Pilot, who was becoming concerned. He therefore contacted the Pilot on board the Blossom Forever and suggested that he alter course away from the Pearl Prosperity.

The Blossom Forever's Pilot agreed and the distance between the ships increased, and the passing manoeuvre was completed safely.

Incident 2

On 15 November 1993, the Liberian tanker Palm Monarch, of 81,282 tonnes deadweight, was overtaking the Australian bulk carrier Iron Shortland, of 107,140 tonnes deadweight. Both vessels were in ballast, each vessel being in excess of 225m in length. The two ships were in the same area of the Great Barrier Reef as the incident of 2 November, but in this case the vessels were north-bound, Iron Shortland bound for the Westem Australian port of Port Hedland and Palm Monarch for the offshore installation of Challis Venture.

Palm Monarch overtook Iron Shortland and the converging courses put the overtaking ship close ahead with both ships on course to pass to the east of Waterwitch Reef. The Master of Iron Shortland expressed concern at the closeness of the other ship and the Pilot altered the ship's course to port, to pass to the west of Waterwitch Reef and any potential risk of collision was averted.

Conclusions

1. Having studied the available information, the Inspector is satisfied that the ships involved in the reported incidents were unnecessarily close, therefore close quarters incidents did occur and that it was reasonable to regard the reports as "incidents" within the meaning of the Regulations.
2. Given that all ships were aware of the close quarter situation and all those on the respective ship's bridges were alert to the potential danger, the risk of collision, had a failure of machinery or steering gear occurred, was remote.
3. Given the situation on the day, there was no compelling reason for ships to overtake in such close proximity.
4. Although communications were established between the vessels involved and the passing manoeuvres discussed, the general level of planning and discussion of projected manoeuvres and courses was inadequate.
5. Given the limits of the water available within the Great Barrier Reef, close quarter situations are bound to occur from time to time. Where vessels' speeds are close and the overtaking manoeuvre prolonged, the overtaken vessel should, in consultation with the overtaking ship, consider reducing engine revolutions to reduce the time of any close proximity.

Summary

Shortly after berthing at West Swanson dock, Melbourne, on 24 September 1993 the 13280 grt container and general cargo vessel "Iron Flinders", on charter to BHP Ltd, suffered a fire in the waste-heat unit (economiser) which is situated in the main engine exhaust trunking at the base of the funnel.

The fire gained sufficient hold for a hydrogen-iron fire to become established, and the internal tube banks of the unit were completely destroyed before the fire was extinguished some eight hours later by the Melbourne Metropolitan Fire Brigade, assisted by the ship's staff and the use of over 8 tonnes of liquid carbon dioxide.

The investigation revealed a combination of several factors which played an important role in causing the fire. The main ones were:

Poor design.

Poor maintenance and operating procedures, due in part to the inadequate provision of on-board instructions and information and, in part, to a lack of instrumentation on the waste-heat unit.

The initial response to the fire revealed considerable uncertainty about, and a lack of specific training for, this kind of fire on the part of the ship's staff. It also revealed some inadequacies in the type of fire-fighting equipment fitted in the vessel.

Conclusions

1. The fire occurred as the result of a combination of factors:

Soot build-up

in order for the fire to have occurred, there must have been a build-up of soot within the unit. The production of soot in the exhaust gas was probably exacerbated by three factors, namely:

1.  
   • The poor performance of the East German fuel valve nozzles,
   • The high consumption of cylinder oil, particularly following the frequent overhaul of pistons, and
   • manoeuvring into port prior to the occurrence of the fire.

The first two of these may have caused fouling of the waste-heat unit over a prolonged period.

The measure employed to assist in the prevention of fouling was the use of Drew Ameroid "LT' soot release powder injected into the bottom of the unit every third day. The effectiveness of this treatment would be reduced if there had been significant oil vapour carry over.

Cleaning

With the inadequate arrangements for thorough inspection of the unit and no provision for water-washing, the quantity of deposits within the more inaccessible areas of the tube nest could have been considerable.

The only method employed for cleaning the unit was that of soot blowing, carried out twice daily. The check-off sheets, however, which were filled in by the Duty Integrated Rating for the two days at sea prior to arriving in Melbourne, that is the 22nd and 23rd September, indicate that the unit was soot blown only once on each of those days, instead of twice as was the routine.

Ignition

Ignition of the accumulated soot could have been caused by either high exhaust gas temperature or by sparks carried over from combustion into the exhaust gas stream. The engine manufacturers, MAN-B&W, have stated in one of their service letters that "oil-wetted soot" may ignite at temperatures as low as 150 Celsius. If the fire, as is most likely, had been smouldering for some considerable period before "Finished with Engines", then the action of shutting down the feedwater circulation would have assisted the fire to gain a hold by stopping the only means by which the heat of combustion was being removed from the tube banks.

2. The lack of information provided by the manufacturers on keeping the tubes clean if part of the unit was to be run in the "dry" condition, is considered to be a contributory factor to the incident. Similarly, the absence of any instructions from the owners, following similar incidents in recent years, emphasising the importance of maintaining waste-heat units in clean condition to minimise the risk of soot fires, is also considered to have been a contributory factor.

No regular monitoring of the degree of fouling of the unit was carried out. The temperatures of gas flow into and out of the unit were not recorded and no instrumentation was fitted for measuring the pressure differential of the gas flow through the unit. Had the remote alarm for "Uptake Temperature High" been connected, an earlier warning of the fire would have been received.

No information was available on board, either in the equipment manuals, owner's orders, or operating instructions concerning the means by which to fight this type of fire. This is considered to have contributed to delays in tackling the fire by the best means in the early stages.

No reports of similar incidents had been passed to their fleet by the vessel's owners, neither was this type of fire covered in the fire training courses undertaken by the ship's staff.

The ship's fire hoses carried in the engine-room, although complying with the requirements of the regulations, were not the most suitable size for use in this situation. Similarly, the nozzles, of East German design, gave a water flow rate and cone angle not well suited to the cooling of surfaces in confined spaces.

Summary

On 28 July 1993, the off-shore supply ship Maersk Runner was employed running out anchors for the mobile drilling unit Atwood Falcon in the Walkley-l field, on the Australian North-West Shelf.

This operation involved the use of a 64 mm diameter working wire, or "chaser wire", which is used to haul the anchors out to the correct position and then to lower the anchors to the seabed.

As a dedicated chaser wire had not been provided, the ship had been using the starboard, 70 mm towing wire. When this towing wire parted, early that morning, a composite chaser wire, made up from a number of wire pendants shackled together, was used.

In the early part of the afternoon, one of the anchors failed to set in the seabed and had to be lifted and repositioned. As Maersk Runner started to lift this anchor, the composite chaser wire parted at a spelter socket, positioned just off the winch. The Mate was struck and thrown by flaying wire, suffering severe head and pelvic wounds, as a result of which he died.

Conclusions

1. The Mate died as a result of injuries received from being struck and thrown by the chain on the end of the port tugger wire and/or the 64 mm diameter chaser wire.
2. The cause of the Mate being struck by the chain/wire was the failure of the spelter socket at the inboard end of a 200m, 64 mm diameter wire pendant, shackled to the ship's working wire and being used as part of a "composite" chaser wire.
3. The spelter socket failed because of the effects of shock overload when the chaser was "homed" on the anchor.
4. The Mate's standing in an exposed position was brought about by the requirements of the particular operation.
5. The operation of using tugger wires on chaser wires, although not a common or usual practice, is necessary when using a composite wire with connecting shackles and sockets within its length.
6. There was no negligence on the part of the Mate, the Master or crew of Maersk Runner.
7. The long, straight stirrup "Crosby" design of spelter socket is not suitable for spooling on to a winch drum of a diameter the size of the normal working drum.
8. Chaser wires should be of one continuous length, to avoid shackles and spelter sockets being spooled on to the drum, and thus also avoid the need for the use of tugger wires.

Summary

On 12 May 1993, the fully laden, 81,248 tonnes deadweight Bahamas flag tanker Oppama Spirit grounded at Gove, Northern Territory. The ship had been manoeuvring to go alongside the berth, when it experienced a total power loss as a result of the seawater cooling system becoming blocked. While under tow to the anchorage, the ship took a sheer and grounded forward.

The ship was pulled clear by one of the tugs after less than one hour and towed to the anchorage without further mishap.

No damage was sustained by the ship and no oil pollution occurred as a result of the grounding.

Conclusions

It is considered that:

1. The blockage of the sea-water cooling system, which resulted in a total power failure, was caused by mud, sand and shells churned up from the seabed and drawn into the system while the engine was being run astern off the berth at Gove.
2. The decision to abort the berthing and tow the ship out to the anchorage was appropriate.
3. The decision not to use an anchor to stop the sheer to port was appropriate.
4. The two tugs stationed at Gove were under-powered for the task in hand and were unable to stop either Oppama Spirit's sheer to port or the grounding.

Summary

The German flag, 35,303 gross tonnes container ship Berlin Express, while proceeding in auto-pilot steering mode through the dredged cut, in the South Channel, Port Phillip Bay, en route to Melbourne, took a rapid sheer to starboard, as a result of which the ship grounded adjacent to South Channel Pile beacon.

Assisted by two tugs from the Port of Melbourne, the ship was successfully refloated 10 hours later, on the next high tide.

The ship suffered no structural damage, and no pollution occurred as a result of the grounding.

Conclusions

Berlin Express grounded as a result of a rapid sheer to starboard brought about by the rudder moving to starboard 20 degrees.

It is considered that:

1. Although an autopilot human input error is the most straightforward and simplest explanation for the movement of the rudder to starboard 20 and cannot be totally discounted, such an error was unlikely.
2. A momentary systems failure, either of the autopilot or the steering gear, was unlikely.
3. Electrical radiation interference of the electronic control systems, from an outside source, was unlikely.
4. Possibly no conceivable malfunction occurred, instead the movement of the rudder to starboard 20 was the response of the autopilot, under the programmed settings, to that particular set of circumstances.
5. The speed of Berlin Express, on entering the cut, was around 18.5 knots.
6. The effect of squat reduced the under-keel clearance as the ship passed between buoys 11 and 12 to 4.78m. As the ship sheered away from the centre line, this would have been reduced to about 2.75m, which would have had an adverse effect on the ship's manoeuvrability.
7. Those on the bridge acted quickly and correctly but were unable to prevent the grounding.
8. At the speed at which Berlin Express was travelling, once the sheer commenced, grounding was unavoidable.
9. Although regulation 339 of the Port of Melbourne Authority (Amendment) Regulations 1988 is no longer considered appropriate, had Berlin Express been proceeding at a much-reduced speed for passage through the dredged cut, the lateral movement to starboard would not have been so great and, therefore, the grounding would have been less likely to happen.
10. The movement of the rudder to starboard 20, the resultant sheer, and hence the grounding, would most probably not have occurred had the steering been conducted manually.