At 0600 on 14 January 1999, the Australian Antarctic research and supply vessel Aurora Australis was on passage from Fremantle to the Antarctic, making good a speed of 11.5 knots on a southwesterly heading. After a 'non-critical' engine room alarm at 0614, a fire alarm occurred. The master, who had just relieved the mate on watch, glimpsed flames on an engine room CCTV monitor. The master sounded the general alarm and stopped the main engine. The ship's staff and the expeditioners went to their fire muster stations, while the duty engineer made for the engine control room. He encountered thick smoke as he entered the engine room but was able to reach the control room, from where he shut down all the running fuel pumps and engine room ventilation supply fans.
At 0627, after receiving reports that all vent flaps were shut, all remote fuel shut off valves operated and all fans stopped, the chief engineer discharged the Halon 1301 fixed fire-extinguishing system. The ship's position was 32º 57.7´ South, 114º 09.7´ East. Seven minutes later, the vessel transmitted a PAN PAN call which, after a number of repeats on different frequencies, was acknowledged by Brisbane Radio.
At 0700, to conserve fuel, the emergency generator was shut down, rendering the vessel 'dead ship'. The master convened a meeting on the bridge to discuss the best method by which to effect a re-entry into the engine room and at 0907, after re-starting the emergency generator, the chief engineer and an IR, backed up by a second team standing by, entered the engine room via the shaft tunnel to assess the situation. They reported that the fire was out then attempted to use a fire hose to cool down the area around the forward end of the starboard main engine, where the fire appeared to have been centred. On pressurising the hose, however, it parted from the coupling. A second hose brought into the engine room also parted from its coupling.
The engine room was vacated while other hoses were tested on deck before another attempt was made to cool down the seat of the fire. It was planned that, because of the heat, teams would enter in 10 minute relays. At 1100 another team entered, cooled the area and checked for hot spots. A bin of smouldering rags was found near the control room and eventually extinguished.
Checking of the engine room for hot spots was continued for the next two hours before the engine room was opened up and venting commenced at 1302. When attempting to start an air compressor, the emergency generator stopped and it was found that a short circuit appeared to have damaged its automatic voltage regulator. The master, concerned about the vessel's situation, with no electrical power and no starting air for the generators or main engines, called P&O Polar and requested that a tug stand by. Soon afterwards, a tug was despatched from Fremantle.
After fitting a spare AVR, which was found also not to work, the engineers supplied excitation for the generator by tapping off the generator's starting batteries and eventually restored power to the vessel. At 2345, the port main engine was started and Aurora Australis, escorted from 0300 by the tug Wambiri, proceeded back to Fremantle at 8 knots, arriving alongside at 1305 on the following day (15 January).
These conclusions identify the different factors contributing to the incident and should not be read as apportioning blame or liability to any particular organisation or individual.
Based on all the evidence available, the following factors are considered to have contributed to the fire:
- The fire was caused by the failure of a flanged joint on the starboard main engine fuel supply pipework, the resulting spray of diesel fuel igniting on the turbochargers or exhaust pipework.
- The flanged joint failed after two of the four screws failed in fatigue, and the other two had worked loose.
- A sheet metal cover, manufactured by local contractors and which left a gap down the side, failed to provide the necessary screening of the failed flanged joint, hence allowing the escaping fuel to reach a source of ignition.
- In the absence of all the physical evidence, the inspector could not reconcile conflicting evidence, given by the engine-builders and ETRS, on whether or not the 8 mm allen screws had, or had not, been overtightened upon last assembly of the pipework.
- The design of the pipework adjacent to the failed flanged joint is such that a torque wrench cannot be applied to all the screws in order to tighten them to the correct torque specified in the Wärtsilä manual.
- Vibration and mis-alignment of the pipework were factors which contributed to the failure of the allen screws.
- The use of allen screws, having no locking arrangement, in the LP fuel system was a further contributing factor.
It is also considered that:
- In general, the response to the fire by the ship's crew and the expeditioners on board was measured, effective, and demonstrated initiative.
- Following an earlier fire on the same vessel, in July 1998, P&O had taken all reasonable measures to reduce the risk of another such engine room fire.
- The use of allen screws without any locking arrangement is becoming an increasingly common contributing factor in fires caused by failures in the LP fuel systems of medium speed diesel engines.
- The maritime industry should note the inspector's concern about the practice of individuals entering a fire zone in the engine room in order to reach the engine control room during the early stages of a fire.
|Date:||14 January 1999||Investigation status:||Completed|
|Location:||Off Sth WA Coast|
|Release date:||17 February 2000||Occurrence category:||Incident|
|Report status:||Final||Highest injury level:||None|
|Type of operation||Research and Antarctic supply vessel (Ice Class)|
|Damage to vessel||Substantial|
|Departure point||Fremantle, WA|