The pilot reported that he commenced the take-off roll on the easterly runway with a quartering crosswind of 5 to 10 kts from the north-east. During the early part of the roll, the aircraft veered to the right before the tail wheel was off the ground. The pilot managed to correct the swing and continued with the take-off. The aircraft swung further right again when the tail wheel was raised. The pilot apparently overcorrected and headed towards the left side of the strip. During this change in direction the left wing lifted, and the right wing struck the ground. The right main gear collapsed during the subsequent slide.

Both occupants evacuated the aircraft from the left side.

The pilot later said that he thought the crosswind swung to a quartering tailwind during the take-off roll.

Following an uneventful flight, the pilot reported that the aircraft touched down normally. After rolling about 80 m the right wing dropped, and the pilot suspected a tyre had failed. But about 50 m further on he heard pronounced scraping noises, and the aircraft slewed right then left before finally coming to a halt about 10 m to the left of the runway.

A subsequent inspection of the aircraft revealed that the fixed landing gear attachment bolts had failed at the countersunk heads. The failure faces contained rust deposits, indicating the bolts had been cracked for some time prior to the accident.

The manufacturer has subsequently changed the design specification to fit standard flat head bolts in place of countersunk bolts.

The crew of the BA146 requested a clearance from the Alice Springs aerodrome controller (ADC) and were cleared via ANGAS and their planned route at flight level (FL) 200. The BA146 taxied late for departure and the crew requested track shortening via the direct route to reduce the delay to the schedule. At FL200 the BA146 would leave controlled airspace at 120 NM Alice Springs (which is 60 NM from Ayers Rock). This was also the point at which the crew of the BA146 expected to commence descent to Ayers Rock.

The ADC had a flight plan for a B737 flight from Ayers Rock to Alice Springs at FL270 and he was aware of the route procedure. The B737 had earlier departed from Alice Springs, but later than scheduled. The ADC surmised that as the aircraft had departed Alice Springs late, then it would depart Ayers Rock late as well. If this was the case, the ADC assessed the BA146 and the B737 would not conflict on the direct track. Also, there was some uncertainty amongst controllers regarding which aircraft were required to track via ANGAS. The ADC understood that aircraft operating at FL200 and below could be cleared on the direct track if requested by the pilot. He recleared the crew of the BA146 to track direct to Ayers Rock at FL200.

Shortly after the BA146 departed Alice Springs the B737 taxied at Ayers Rock and requested a clearance from sector control. The clearance request for the B737 was relayed via flight service in accordance with air traffic service (ATS) procedures. Ayers Rock aerodrome is located outside controlled airspace (OCTA) and the sector responsible for the area does not have radio facilities which would enable direct speech with aircraft on the ground. The B737 crew was instructed to track from Ayers Rock to Alice Springs and to maintain FL190 due to the BA146 at FL200. The crew of the B737 was provided with traffic information on the BA146 plus other pertinent aircraft. After departing, the crew of the B737 reported to flight service that the aircraft was maintaining FL190.

Following discussion between the ADC and the sector controller, it was agreed that the B737 would be transferred to the ADC, from flight service, to obtain a clearance into controlled airspace (CTA). The sector controller concurred with the ADC's intention to clear the B737 into CTA on climb to FL270 after separating the aircraft from the BA146. The crew of the BA146 reported at 120 NM Alice Springs. At this point the BA146 was leaving CTA for OCTA, and the ADC was no longer required to provide separation from other aircraft. The ADC issued traffic information on the B737 including the departure time from Ayers Rock and that it was tracking in the opposite direction on climb to FL270. He then instructed the crew to contact flight service. The ADC did not advise the crew of the BA146 that the B737 had been instructed to expect a clearance at FL190 or that aircraft's estimate for Alice Springs.

The crew of the BA146 changed to the flight service frequency and reported their position as 56 NM east of Ayers Rock, left FL200 and estimating the aerodrome at 0344 UTC. Prior to leaving CTA the crew of an IFR category aircraft is required to report position on the flight service frequency. This is to ensure that other aircraft, which may be flying in close proximity of the boundary between CTA and OCTA, can take action to avoid the aircraft leaving CTA. Alternatively, the change to the flight service frequency enables the crews of any potentially conflicting aircraft to converse and arrange their own separation. Also, the crew of an aircraft which intends to change level OCTA is required to report the intended change to flight service one minute prior to commencing the change of level. Again, this is to ensure that other aircraft which may conflict have time to avoid or to arrange separation from the aircraft changing level. The crew of the BA146 did not complete either of these actions before leaving CTA or descending from FL200.

The BA146 crew advised flight service that they had traffic information on the B737 on climb to FL270. Flight service immediately advised the crew that the B737 was on climb to amended level FL190. The BA146 crew asked whether the B737 had departed Ayers Rock. Flight service advised the departure time of the B737, that the aircraft was maintaining FL190 and that the B737 crew was operating on the flight service frequency. The BA146 crew then transmitted that they had left FL180. The two crews subsequently advised their respective distances from Ayers Rock which indicated they were 10 NM apart and yet to pass each other. The crew of the BA146 had descended their aircraft through the level of the B737 prior to establishing the relative positions of the two aircraft.

There was no breakdown in separation.

ANALYSIS

After falling behind schedule, the crew of the BA146 requested air traffic control approval to track direct to Ayers Rock in an endeavour to make up time. This request was contrary to the preferred routing detailed in ERSA. The crew were aware that the diversion route was implemented to minimise the potential for conflict between aircraft operating between Alice Springs and Ayers Rock. However, they felt that the commercial interests of the airline would be better served in regaining the schedule. They did not give due consideration to the flight safety aspects inherent in their decision.

The ADC was unsure of the correct procedure regarding which flights could operate on the direct track from Alice Springs to Ayers Rock. This was despite the fact that the ERSA stated that the preferred route for aircraft operating between Alice Springs and Ayers Rock was via ANGAS. He assessed that the B737 would not conflict with the BA146, but he had no means to ensure that this remained the situation, especially during the period of each aircraft's flight outside controlled airspace. Use of the diversion route via ANGAS by the BA146 would have ensured that the two aircraft did not conflict outside controlled airspace.

The ADC did not provide accurate traffic information to the crew of the BA146 before the aircraft left CTA. The advice that the B737 was to climb to FL270 may have led the crew of the BA146 to expect that the B737 would be above their level before they left CTA. Also, the ADC did not advise the crew of the BA146 of the B737's estimate for Alice Springs. Provision of this information would have enabled the crew of the BA146 to estimate the time they would pass the B737 which would have assisted them in assessing the potential for conflict OCTA.

The inaccurate traffic information provided by the ADC about the B737 may have lulled the crew of the BA146 into thinking that there was no likelihood of conflict OCTA. Consequently, they may not have been as vigilant in their adherence to procedures as usual. The lack of the report by the BA146 crew, on the flight service frequency, at least one minute prior to commencing descent was a failed defence for the safe conduct of the flight. A second failed defence was the lack of the transmission of a position report on the flight service frequency prior to leaving CTA. Had the crew of the BA146 made either of these reports there was every possibility that the crew of the B737 would have transmitted their position details. This would have alerted the crew of the BA146 to the fact that the two aircraft may conflict once their aircraft descended, and the two crews could have arranged appropriate separation.

SIGNIFICANT FACTORS

1. The crew of the BA146 requested the direct track contrary to the preferred routing instructions in the ERSA.
2. The ADC was not sure of the requirements relating to aircraft tracking to/from Ayers Rock.
3. The ADC did not provide adequate traffic information to the crew of the BA146.
4. The crew of the BA146 did not comply with radio reporting procedures detailed in AIP

SAFETY ACTION

Local safety action

1. Airservices Australia issued a temporary local instruction to air traffic controllers instructing that aircraft operating between Alice Springs and Ayers Rock at altitudes above 10,000 ft, and intending to land at Ayers Rock, are to be cleared via ANGAS unless weather conditions preclude the use of the track.
2. The operator of the BA146 issued two notices to pilots. The first notice clarified the radio procedures to be used by pilots when operating in CTA prior to descending OCTA. The second notice instructed pilots that they must track via ANGAS (unless ATC requires alternative routing) and also reminded them of the requirement to report to flight service prior to leaving CTA.

FACTUAL INFORMATION

A pilot undergoing a flight check was conducting instrument flight rules (IFR) Airwork in a Partenavia (P.68C) at Darwin. The pilot requested approval from the approach radar (APP) controller to conduct a runway 11 locator approach. This would entail the P.68C flying a tear drop pattern from overhead the aerodrome, then tracking out to the northwest with a right turn to return to the aerodrome. The initial holding altitude for the procedure is 3,000 ft, with the aircraft descending while flying the tear drop pattern. The APP controller was expecting a Boeing 767 (B767) to depart shortly from runway 29 en route to Singapore and on climb to flight level (FL) 310. The outbound track for the B767 was 294 degrees. To maintain separation between the two aircraft, the APP controller decided to assign separation responsibility to the pilot of the P.68C. The APP controller asked the pilot to report sighting the B767 which was taxiing for departure, and whether he could maintain separation with the B767. The pilot of the P.68C stated that he could see, and would maintain his own separation with, the B767. The pilot of the P.68C explained to the APP controller that he had a check captain on board and that the latter would maintain visual contact with the departing B767. The APP controller instructed the pilot of the P.68C to make the locator approach and to maintain 3,000 ft.

The Manual of Air Traffic Services (MATS) states that visual separation can be used "when a pilot, at or below 10,000 ft reports sighting another aircraft and is instructed to maintain visual separation from it." The Aeronautical Information Publication (AIP) which details the rules and procedures applicable to pilots contains a corresponding reference to the visual separation standard. Neither the MATS nor the AIP detailed a minimum lateral or vertical separation distance to be maintained by the pilot in the application of the visual separation procedure. The AIP further states that if a pilot is given responsibility for separation from another aircraft, he is also responsible for the provision of wake turbulence separation. The crew of the B767 were not aware of the P.68C. The APP controller was not required to advise them that the pilot of the P.68C had been assigned visual separation responsibility. It was early in the evening and consequently the crew of the B767 were looking into the setting sun as the aircraft departed. After becoming airborne the crew noted a target on the aircraft's Traffic Alert and Collision Avoidance System (TCAS). The TCAS indicated that there was an aircraft 1,800 ft higher than the B767 and on their intended track.

The crew of the B767 attempted to sight the other aircraft but were unable to see because of the glare from the sun. The TCAS indicated the vertical separation was 800 ft and rapidly closing when the crew were able to sight the aircraft just off to the right of the track of the B767. The B767 crew decided that to comply with their clearance, which would have required a right turn towards the other aircraft, would cause the two aircraft to conflict. Consequently, the B767 crew maintained runway heading to avoid the P.68C. The check captain in the P.68C had observed the departing B767 and believed that there was sufficient horizontal separation between the two aircraft. Both the pilot and the check captain of the P.68C were aware of the intended track of the B767. The B767 crew contacted the APP controller and requested traffic information on the P.68C. The APP controller advised them that the pilot of the P.68C had the B767 in sight, was maintaining visual separation and was at 3,000 ft. Subsequently, after passing the P.68C and leaving 3,000 ft, the crew of the B767 turned their aircraft to intercept the outbound track. As there are no radar recording facilities at Darwin the actual minimum horizontal and vertical separation between the two aircraft could not be ascertained. The APP controller's use of the visual separation standard was in accordance with the MATS which authorised the transfer of separation responsibility to the pilot under certain circumstances. As the visual separation procedure has no minimum separation distances for use by the pilot in the application of the procedure, and the check pilot in the P.68C had the B767 in sight, there was no breakdown in separation.

ANALYSIS

The provision of traffic information to the crew of the B767 was not required. However, issuance of traffic information on the flight of the P.68C prior to departure would have reduced the element of surprise when the crew of the B767 became aware of the other aircraft. The B767 crew's concern with the location of the P.68C was understandable given that they were not aware that the crew of the other aircraft had them in sight and were maintaining visual separation. The installation of TCAS to aircraft enables flight crews to monitor the proximity of other aircraft with an operable transponder. Consequently, crews with access to TCAS are usually more aware of traffic in proximity to their aircraft. There may be benefit in providing traffic information to crews in similar situations to avoid unnecessary deviations from the cleared route and/or altitude. The check captain continually observed the B767 while the pilot of the P.68C flew the runway 11 locator approach. The amount of horizontal or vertical separation required when using the visual separation procedure is at the discretion of the responsible pilot. Neither the MATS nor the AIP details a minimum horizontal or vertical separation distance or provides any guidance for what is expected or is considered safe. Currently, the minimum separation distance is subjective and conditional upon the training and experience of the pilot using the procedure. Publication of a minimum horizontal or vertical distance to be maintained between aircraft during the application of visual separation would enable flight crews and air traffic controllers to assess whether the procedure was being applied correctly.

SIGNIFICANT FACTORS

1. The pilot of the P.68C accepted responsibility for visual separation from the B767.

2. The visual separation procedure did not provide a minimum horizontal or vertical separation distance to assist pilots in the safe application of the procedure.

3. The crew of the B767 were not provided with traffic information on the P.68C.

SAFETY ACTION

The Bureau of Air Safety Investigation is evaluating the provision of traffic information by air traffic control, and guidance for the application of the visual separation procedure. Any forthcoming recommendations will be published in the Quarterly Safety Deficiency Report.

Factual Information

The flight was conducted from a grass strip about 1,000 m long oriented 060/240 degrees M and operated by a parachuting club. The wind was from the south-west at 10-20 kts.

Witnesses reported that, as the pilot commenced to taxi the aircraft from its hangar to the loading position, the propeller struck the ground. It was not clear whether the impact was sufficient to stop the rotation of the propeller. The aircraft was pulled clear of the area and the propeller examined by the pilot and the jumpmaster. The jumpmaster indicated that the blades appeared to have retained their shape and were intact. However, at his suggestion, the pilot flew a circuit to check the aircraft's performance. The pilot indicated that the aircraft performed normally. Neither the pilot in command nor the jumpmaster were qualified to inspect the propeller in accordance with Civil Aviation Order Part 106 which details the inspections to be conducted following a propeller strike.

Four parachutists then boarded the aircraft. They reported that the ground roll required for the aircraft to take off seemed to be longer, and the aircraft climbed at a shallower angle, than they had previously experienced, however, the engine sounded normal. At what they thought was about 200 ft above ground level, they heard the pilot say that the engine had lost power. He then commenced to turn the aircraft to the left. When the aircraft was aligned with the strip, one of the parachutists called for the pilot to lower the flaps.

Ground witnesses saw the aircraft turn left and approach the strip downwind. The aircraft touched down about halfway along the strip but immediately became airborne again and commenced a shallow climb with a nose-high attitude. The flaps were observed to be in the fully extended position and the aircraft was rolling left and then right. About 300 m beyond the end of the strip, the aircraft struck powerlines 8 m above ground level, then collided with the ground at a steep nose-down angle and was destroyed by impact forces.

The aircraft was fuelled to capacity a few weeks before the accident. There was no evidence to suggest that it had been flown or defueled in the intervening period. The weight of the aircraft at take-off was calculated to be approximately 1,198 kg. This compares with its maximum take-off weight of 1,043 kg.

Examination of the propeller revealed damage to both blades. While some damage appeared to be consistent with the propeller strike during taxi, other damage may have been caused by impact with the ground following the wire strike. Examination of the flaps confirmed that they were fully extended at impact. There was no evidence of any other condition which might have affected the operation of the aircraft.

The aircraft was fitted with only the pilot's seat as it is common practice in parachute operations to remove all other seating. While the pilot's seat was fitted with a lap-sash safety harness, there was no evidence of the presence of safety harnesses or restraints available to the parachutists for the accident flight. Some of the parachutists on board the aircraft advised that it is common practice for parachutists not to wear safety harnesses because of their potential to hinder egress from the aircraft.

The pilot was unable to be interviewed during the course of the investigation. As far as could be determined, his flying experience in the few months before the accident was limited to the conduct of parachute dropping flights from the airstrip on some weekends. His experience on the aircraft type could not be established.

Analysis

There was insufficient evidence to draw any conclusion concerning what affect, if any, the propeller ground strike had on the performance of the aircraft. However, the take-off and climb performance of the aircraft with the parachutists on board would have been significantly less than that which the pilot had experienced during the check flight which had been conducted a short time earlier. The difference in performance may have been sufficient to lead him to believe that the engine had lost power.

The attempt to land downwind in the prevailing conditions contributed to a higher-than-normal approach profile and approach speed.  Similarly, the subsequent attempt to go around with the flaps fully extended resulted in a shallow climb gradient. The observed rolling motions of the aircraft were consistent with it operating close to the stalling speed. The nose-high attitude of the aircraft may have prevented the pilot in command from seeing the power lines.

Prior to departing Narrabri, the pilot obtained a weather forecast for his destination, which indicated the surface wind would be from the southwest at 15 kts. The flight was uneventful and on entering the circuit at Aeropelican the pilot noted that the windsock indicated a northerly wind and elected to land on runway 07.

The pilot later reported that the aircraft crossed the threshold at 80 kts, with full flap selected, and touched down about 250 metres into the 880 metre strip. About halfway along the runway, with the airspeed still indicating 55 kts, the pilot realised there was a tailwind component. He considered going round but elected to continue with the landing. Heavy pressure was applied to the rudder pedals, without operating the brakes effectively. As a result, the aircraft overran the runway end and collided with the airport fence. The four occupants managed to evacuate the aircraft, but the pilot suffered superficial injuries.

An Australian-registered B747 was maintaining flight level (FL) 350 on a flight from Sydney to Hong Kong and was approaching the Zamboanga very high frequency omni-directional radio range beacon (VOR). The crew received a traffic alert and collision avoidance system (TCAS) traffic information advice on an unidentified aircraft at the same level, crossing from right to left and also converging on the Zamboanga VOR. The crew of the B747 initiated a right turn that ensured that their aircraft would pass behind the other aircraft. There was a breakdown of separation.

Investigation by the Bureau in conjunction with the Air Transportation Office of the Republic of the Philippines, established that the other aircraft was foreign-registered and that the crew of that aircraft had been instructed by Manila air traffic control to climb to FL 370 for separation with the Australian aircraft. The instruction had been correctly acknowledged by the crew at the second attempt. The instruction included an authorisation to descend back to FL 350 when 5 minutes past the Zamboanga VOR.

Manila air traffic control had taken all necessary action to ensure positive separation of the two aircraft. The reason why the foreign-registered aircraft was not at FL 370 when passing the Zamboanga VOR could not to be determined.

The crossing point of the two air routes at Zamboanga was not under radar surveillance at the time of the occurrence. However, as part of the improvement to the air traffic system by the Government of the Philippines, a new radar head was commissioned late in 1996 and now provides coverage in this area.

The captain reported that during the landing roll, while decelerating through approximately 40 knots, the nosewheel entered a soft patch on the dirt strip. The aircraft slewed slightly to the right and the nosewheel broke through the rolled crusted surface of the strip. This resulted in the left side of the aircraft being sprayed with large stones, causing damage to the engine, left flap and left wing. The crew (captain and co-pilot) inspected the damage, and the captain elected to fly the aircraft back to Essendon for repairs.

After arrival at Essendon, it became evident that the damage was more extensive than it appeared when the aircraft was inspected by the crew immediately after the occurrence.

Some of the communication problems appeared to be attributable to the similarity of several of the aircraft callsigns. It is probable that the use of HF communications with some aircraft also contributed to the controllers' communication difficulties.

The pilot of VH-PJM was enroute Forbes to Brewarrina at 6,000 ft initially estimating Nyngan at 0941 EST but later revised the estimate to 0931. The pilot of VH-XMM was enroute Walgett to Hamilton at 6,000 ft initially estimating Tottenham at 0910 but revised the estimate twice, the last being 0921. Based on the original flight plan/ flight progress strip estimates for both aircraft, the time of passing in the Tottenham area would have been 0905 with PJM approximately 45 NM south-south-east of XMM.

The pilots revised their estimates enroute because XMM encountered a headwind component whereas PJM encountered a tailwind component. However, by the time the Flight Service operator had passed traffic information to both pilots, both aircraft had passed within 5 NM of each other at 6,000 ft.

The Flight Service operator advised that at the time of the breakdown in separation his traffic workload was fairly high. He failed to realise that the revised estimates placed both aircraft in the Tottenham area at approximately the same time.