Cargo handlers unloading an A330 aircraft in Perth reported a burning smell coming from part of the cargo. When the load was disassembled, evidence of a possible cargo fire was found. Three wooden load spreaders from two different pallets showed signs of scorching, and one aluminium pallet had a 100 mm X 120 mm hole worn through its base, with fused metal around the edges of the hole. Another pallet showed signs of scorching. The affected areas were still hot to touch.

The 16,000-kg load, comprising 16 wooden (pine) crates measuring 1500 mm X 1000 mm X 300 mm and several smaller items, and was divided evenly between four 3000 mm X 2000 mm X 10 mm aluminium pallets. Each of the crates was fitted with three 100 mm X 50 mm X 1000 mm wooden runners which were oriented longitudinally when loaded on the pallets. Each pallet load was positioned on load spreaders which were located transversely. Two pallets were loaded in the rear and two in the forward cargo hold of the A330. The load was prepared in the United Kingdom and transhipped in Kuala Lumpur. Nothing abnormal was reported during transhipment.

The size of the spreaders used, and the damage sustained on the four pallets differed significantly. The load spreaders on all four pallets exhibited evidence of movement between the spreaders and the pallets. Only the two rear pallets exhibited scorching with the rearmost pallet being the most damaged. Four 100 mm X 50 mm spreaders were used on the rearmost pallet.  The spreader located closest to the rear of the pallet was worn down to half its original thickness in the vicinity of the hole. A 300-mm length centred on the hole exhibited signs of significant scorching. No scorching was evident on the other three spreaders. On the next pallet four 130 mm X 20 mm spreaders were used. The front and rear spreaders and their corresponding positions on the pallet exhibited signs of scorching, but only on the starboard side. No scorching was evident on the other two spreaders. Four spreaders, each measuring 130 mm X 50 mm, were used on each of the two forward pallets. No scorching was evident. Witness marks on the pallets indicate that most of the movement between the spreaders and the pallets was longitudinal rather than lateral.

The cargo handlers reported that the load was secured to the pallets by nets and the pallets were secured in the aircraft by tie-down straps. They also reported that the straps appeared to be adequately tensioned. Actual tie-down arrangements could not be checked as the straps were removed prior to discovery of the scorching and their position was not noted. There was no evidence of damage from the aircraft's roller conveyor wheels.

An analysis of the timber used in the damaged spreaders indicated it was European Silver Fir. The samples contained no resin, which was considered unusual. The absence of resin reduced the flammability of the timber.

It is evident that the damage and scorching were the result of friction-generated heat caused by movement between the spreaders and the pallet. The movement apparently did not result in fire as any flame would have spread rapidly to the pine crates resting on the pallets. There was no evidence of this. The cargo hold overheat warning system did not activate, and the cargo hold fire suppression system was not used.

No additional evidence is available to determine why the load was able to move in the way that it did and cause the scorching.

The Bureau of Air Safety Investigation is not aware of any other similar occurrences.

Safety Action

When the timber had been identified, the Bureau of Air Safety Investigation issued a safety advisory notice, SAN960053, informing the Civil Aviation Safety Authority and operators of high-capacity regular public transport aircraft of the circumstances of the occurrence.

The aircraft with an instructor and student pilot on board, bounced heavily on landing. The instructor advised that he did not react quick enough to prevent the aircraft touching down again on the nose landing gear. The nose landing gear then broke off and the aircraft nosed over onto its back. Both occupants exited the aircraft safely.

A British Aerospace 146 (BA146) was established in the holding pattern at YANGO, 60 NM NNW of Sydney, maintaining Flight Level 190 (FL190). A Boeing 737 (B737) entered the holding pattern at FL200, and was given traffic information on the BA146, which was directly below.

Holding was cancelled for the BA146, which was cleared to descend to FL180. The Arrivals North radar controller subsequently observed the BA146 height readout indicating FL200, and was about to query the level of the aircraft, when the crew of the BA146 advised the aircraft had climbed to FL200 but was now on descent to FL190. The lateral separation between the two aircraft was reduced to 1.53 NM, with a vertical separation of 800 ft. The minimum required separation was 5 NM laterally, or 1,000 ft vertically.

The Captain of the BA146 subsequently reported that the aircraft was being controlled by the autopilot at the time of the occurrence. FL180 had been correctly set on the altitude controller, but the aircraft had inadvertently climbed without warning. The captain, who was the non-flying pilot, had earlier given permission to permit three children to stand near the flight deck door. As the aircraft commenced descent to FL180 he had turned around briefly to say goodbye. When he turned back, he saw the aircraft was climbing at about 1,500 fpm, approaching FL200. He immediately instructed the co-pilot to descend to FL190, and alerted ATC.

The flying instructor was conducting dual training for the issue of a retractable undercarriage endorsement. During a touch and go, before take-off power was reapplied, the student momentarily selected the landing gear up while the aircraft was rolling at about 30 knots on the runway. The nosewheel managed to retract, the propeller struck the runway, and the nose landing gear doors were damaged. The aircraft came to rest on the runway.

During the landing roll the instructor retracted the flaps and opened the engine cowl flaps. He saw the student move to select the gear up but before being able to prevent it, the student had moved the selector to 'up', realised his mistake and immediately selected 'down' again.

The aircraft departed Melbourne with an unserviceable auxiliary power unit (APU). During climb the left engine driven generator also failed. As the aircraft was left with only one serviceable generator, (the right engine driven generator) the flight returned to Melbourne.

These problems had been intermittently occurring for a couple of days. After extensive investigation maintenance found that the APU start switch was unserviceable. After the switch was replaced, the APU operated satisfactorily. The left engine driven generator and electrical contractor were subsequently confirmed to be unserviceable and were replaced. Subsequent operations have been satisfactory.

FACTUAL INFORMATION

A Boeing 737 was maintaining FL200 while enroute from Sydney to Canberra via Wollongong and a Beech 1900 was maintaining FL190 while enroute from Cooma to Sydney via Shelleys and Bindook. A Saab 340 was enroute from Sydney to Canberra at FL200 tracking via PAGER and Shelleys. The crews of all three aircraft were operating on the same air traffic control frequency when the crew of the Beech 1900 requested climb to FL200 due to turbulence.

The air traffic controller had previously assessed that the Beech 1900 would be a possible restriction for the descent of the Boeing 737 into Canberra. However, when the pilot of the Beech 1900 requested climb to avoid the turbulence, the controller forgot this potential conflict.

There was a route limitation of not above FL190 without prior coordination with the northern adjacent sector. The controller subsequently assessed that there were no restrictions for the climb and that the Beech 1900 would pass safely in front of the Saab 340. Consequently, the controller coordinated a clearance for the Beech 1900 to climb to FL200 and instructed the pilot to climb to that level. As the pilot of the Beech 1900 commenced climb to FL200 the horizontal distance between that aircraft and the Boeing 737 was approximately 6 NM. The controller then recognised the potential conflict between the Boeing 737 and the Beech 1900 and immediately instructed the pilot of the latter aircraft to maintain FL190. However, the pilot of the Beech 1900 had already left FL190 on climb and was unable to descend back to the original level before the aircraft reached FL195. The horizontal distance between the Beech 1900 and the Boeing 737 at the time was 3 NM. The required separation was 5 NM horizontally or 1000 ft vertically.

The controller instructed the crew of the Boeing 737 to turn left onto a heading of 180 degrees to maintain horizontal separation. However, before turning the aircraft, the crew reported that they had the Beech 1900 in sight. The controller provided an update of the position of the Beech 1900 as the aircraft passed 4 NM abeam. The pilot of the Beech 1900 reported at FL190, and the crew of the Boeing 737 reported turning left onto the radar heading. There was a breakdown in separation.

ANALYSIS

The controller could offer no reason for not including the Boeing 737 in his considerations. Possibly, he became focused on providing relief from the turbulence for the Beech 1900 to the detriment of his overall control. Also, the need to ensure that coordination with the adjacent sector was completed prior to approving the climb for the Beech 1900 may have diverted his attention.

SIGNIFICANT FACTORS

The controller forgot the relative positions of the aircraft when he approved the pilot of the Beech 1900 to climb to FL200.

FACTUAL INFORMATION

An Airbus A320 and a B737 departed from Melbourne enroute to Sydney approximately two minutes apart. The A320 departed first and the crew had planned to cruise at FL370, while the crew of the B737 had planned at FL330. Both crews were cleared by air traffic control to climb to their respective levels. Airspace procedures for Instrument Flight Rules (IFR) flights operating in Class C airspace below 10,000 ft require aircraft to operate not above 250 kts indicated airspeed (IAS). Air traffic control may amend or cancel the speed restriction if the reduced speed is not required.

After the two aircraft were airborne, the Departures North (DEPN) radar controller cancelled the speed restriction by instructing both crews to climb at their desired speed. The DEPN controller did not notify any other air traffic control position that the speed restriction had been cancelled. Both aircraft transferred to the Inner North radar controller and, soon after, the crews were each instructed to reach FL330 by 80 NM Melbourne. Both crews acknowledged and readback the level requirement. The Inner North position had assumed control responsibility for the Inner West radar position due to the low level of traffic. This was normal practice when traffic levels at a position reduced such that the responsibility for the position could be consolidated with another control position. The controller was controlling one additional aircraft that was not normally the responsibility of the Inner North position. Horizontal separation between the aircraft was approximately 12 NM with the ground speed of the A320 approximately 10 kts faster than the groundspeed of the B737.

The air traffic control radar display provides a readout of aircraft ground speed which is used by controllers to assist in separating aircraft. During the next ten minutes the Inner North radar controller became involved in coordinating the separation and sequencing of a number of other aircraft and the horizontal distance between the A320 and the B737 reduced to 8 NM. The groundspeed of the B737 had increased and was approximately 100 kts faster than the groundspeed of the A320. A controller from another radar position contacted the Inner North radar controller and queried him with regard to the high rate of closure between the two aircraft and the potential to lose horizontal separation. The Inner North radar controller instructed the crew of the B737 to turn left 30 degrees and to report the new heading. The B737 crew reported that the heading would be 360 degrees and asked the controller for the type of the aircraft ahead of them and, shortly after, for the level of that aircraft. The horizontal distance between the two aircraft was then 5 NM, with the groundspeed of the B737 just over 100 kts faster than the groundspeed of the A320.

The Inner North radar controller advised the crew the aircraft ahead was an A320 at FL310. The controller was required to acknowledge and transmit instructions to a number of aircraft before being able to return his attention to the A320 and the B737. The horizontal distance between the aircraft had reduced to 4 NM, with vertical separation of 700 ft and the groundspeed of the following B737 approximately 90 kts faster than the groundspeed of the A320. The crew of the B737 reported sighting the A320 and vertical separation of 2,000 ft was achieved after another two minutes. The controller cancelled the radar heading for the crew of the B737 and instructed them to rejoin their planned route. There was a breakdown in separation.

ANALYSIS

The cancellation of the speed restriction for aircraft below 10,000 ft by the Departures radar controller was not in itself a significant factor in the incident. However, the fact that the Departures radar controller did not advise the next control position that he had cancelled the restriction was significant, as the Inner North radar controller was not alerted to the possibility of a ground speed differential between the two aircraft. Had the Inner North radar controller known of the cancellation of the speed requirement, he may have more closely monitored the progress of the two aircraft.

A significant portion of the inner North radar controller's time was spent co-ordinating with another control position and this would appear to have diverted him from adequately monitoring aircraft separation on the radar display. The inquiry by another controller alerted the Inner North radar controller to the proximity of the two aircraft and the high rate of closure due to the different groundspeeds. The Inner North radar controller became aware that the minimum horizontal and vertical separation was not going to be maintained and instructed the crew of the B737 to turn the aircraft away from the A320.

SIGNIFICANT FACTORS

1. The Departures North radar controller did not advise the Inner North radar controller that the speed restriction on the two aircraft had been cancelled.

2. The Inner North radar controller was distracted by coordination with another control position and did not adequately monitor the progress of the A320 and the B737.

3. Action by another controller alerted the Inner North radar controller to the proximity of the A320 and B737 and the significant difference in the groundspeeds.

SAFETY ACTION

The Bureau of Air Safety Investigation is evaluating aspects of separation assurance techniques within air traffic control. The details of this occurrence will be used to support SADN 960051. Any forthcoming recommendations will be published in the Quarterly Safety Deficiency Report.

FACTUAL INFORMATION

A B200 had departed Amberley, Qld for Rockhampton, Qld and was tracking from Kilcoy to Gayndah on climb to flight level (FL) 200.  This level was an initial air traffic control restriction pending further climb approval from Brisbane Sector 3B.  The pilot contacted sector control prior to Kilcoy and was re-cleared from the aircraft's present position direct to Gayndah.  This instruction had the effect of establishing the aircraft on a track slightly to the west of the Kilcoy direct Gayndah track.

The sector controller had a foreign registered A330 in conflict with the B200 and instigated a step climb procedure until the pilot of the B200 was assigned his preferred level, FL280.

A B727 had departed Brisbane, Qld for Cairns, Qld and was on climb to the standard intermediate level, FL200.  It was on a track that included the leg Kilcoy to Gayndah.  The crew contacted Brisbane Sector 3B and the controller issued a climb instruction to the flight planned level, FL350.  The controller also estimated that the B727 would pass the B200 just prior to Gayndah and, at 1129 EST, asked the crew of the B727 if they could reach FL290 by 80 NM north of Brisbane; a requirement that, if accepted, would have guaranteed separation between the two aircraft.

The B727 crew replied that they could not make the requirement, and the controller acknowledged, informing the crew that a radar vector for separation may become necessary.

Approximately four minutes later, the controller's attention was transferred to the task of processing several aircraft in the Maroochydore area that were in conflict and required separating.

At approximately 1138, as the B727 climbed through FL270, separation with the preceding, slower B200 broke down.  This was not immediately noticed by the controller.  However, when making a subsequent periodic scan of his traffic, he did notice the conflict and initiated a radar vector for the B727. This action re-established radar separation at 1140.

The B727 passed 2.3 NM to the east of the B200 while passing through that aircraft's level. The required standard was 5 NM by radar.

ANALYSIS

The controller had recognised the conflict between the two aircraft and had correctly estimated the point at which radar separation would break down. He attempted to establish positive procedural separation but, when the crew of the B727 were unable to meet the proposed requirement, elected to radar monitor the aircraft knowing that both vertical and horizontal separation were likely to be lost eventually.

Alternative procedural standards were available to establish positive separation, but these options were not taken by the controller.  When he became task oriented to the situation developing to the south-east of his area of responsibility, the time available to continue radar monitoring the B727 and B200 in the northern part of his airspace was severely reduced.

Consequently, the controller's traffic scan did not detect the proximity of the two aircraft until a breakdown in separation had occurred.

SIGNIFICANT FACTOR

The sector controller did not apply appropriate separation assurance techniques.

SAFETY ACTION

The Bureau of Air Safety Investigation is evaluating aspects of separation assurance techniques within air traffic control. Any forthcoming recommendations will be published in the Quarterly Safety Deficiency Report.

As the engine had missed twice on the previous leg, the pilot carried out an engine check before departure. All indications were normal. About four minutes after departure, the engine became progressively rougher. The aircraft turned back to Gove and landed safely.

The number two piston was severely eroded, possibly due to detonation.

The flight was planned from Maroochydore to Griffith with a refuelling stop at Nyngan. On arrival at Nyngan the refuelling agent could not be contacted so the pilot diverted to Dubbo to refuel. Prior to departure from Dubbo, the pilot estimated that he would arrive at Griffith at last light. He was not night VFR rated. The aircraft ultimately arrived at Griffith at about last light. Conditions were dark and runway lights were required for landing. The pilot made several circuits of the aerodrome while he attempted to turn on the pilot activated aerodrome lighting (PAL) system.

The pilot reported that he made two attempts to turn on the PAL but without success. He then made an attempt to land without the aid of runway lighting. After touchdown he realised he had landed well beyond the threshold and would not be able to stop by the end of the runway. He applied power and attempted to go around but the aircraft stalled and hit the ground in a left wing low attitude after which it cartwheeled to a stop. Although the aircraft was substantially damaged and there was still a significant amount of fuel in the tanks, there was no post impact fire.