Report not released due to no IP comment from the Indonesian agency on this joint investigation report. ASOR hard copy held on occurrence file. Report not to be released without the approval of the Director or a DD Signed Alan L Stray Deputy Director Investigations 6 July 1998

It was reported that the solo student pilot made a normal approach to runway 36 but elected to start a go-around because the aircraft was not aligned with the runway centreline. As the pilot applied full power the nose yawed to the left. At this point, and contrary to training, the pilot selected the flaps to the take-off position. The aircraft descended, touching down to the left of the strip. It then ran off the runway and flight strip and into a rough area suffering extensive damage in the process.

FACTUAL INFORMATION

The pilot of a Shorts (SD3-60) aircraft inbound to Mackay aerodrome from Proserpine at 5,000 ft, requested runway 14. Runway 32 was the duty runway being broadcast on the automatic terminal information service. The aerodrome controller (ADC) instructed the pilot to descend to 2,000 ft and advised him that the downwind on runway 14 was 13 kts. This exceeded the aircraft's downwind limitation of 10 kts and the pilot requested to be advised of the wind when the aircraft was closer to the aerodrome. The ADC advised the pilot to expect runway 14 and to report at 10 NM.

The pilot of a Piper (PA28) aircraft, having completed a session of circuit training, requested further circuits on runway 32. After the PA28 became airborne, the instructor in the aircraft discussed the conduct of a simulated radio failure with the ADC. The ADC cleared the pilot of the PA28 to make a touch and go, and instructed him to make a right circuit. The instructor then commenced the radio failure exercise.

The pilot of a Hughes (H369HS) helicopter requested a clearance to transit the control zone. The H369HS had departed Hay Point (8 NM south-east of Mackay) for Walkerston (7 NM west of Mackay). The pilot was instructed to track as requested at not above 1,000 ft. The intended track of the helicopter passed approximately 2.5 NM to the south-south-west of the aerodrome.

The pilot of the SD3-60 reported at 10 NM and the ADC advised him that runway 14 was not available due to circuit traffic and to track for a left downwind to runway 32. This required the aircraft to be tracked to the south-west of the aerodrome. The pilot maintained 2,000 ft and tracked for left downwind. The PA28 was on right base runway 32 and was approved to land using light signals in accordance with the "no radio" procedures. The tower co-ordinator operated the signal lamp to assist the ADC. After landing, the instructor of the PA28 advised the ADC that the training was finished.

The SD3-60 entered left downwind, and the ADC cleared the pilot to make a visual approach. The issue of a visual approach clearance enabled the pilot to descend, as required, from the last acknowledged level of 2,000 ft to the threshold of the assigned runway. The track and descent of the SD3-60 passed through the level, and close to the track, of the H369HS. While two or more aircraft are in conflict, air traffic control was required to ensure that one of the forms of separation was being applied. Separation can be applied by the use of either radar, visual, vertical, lateral or longitudinal standards. There was no radar display in the tower and the ADC was thus required to use one of the other forms of separation. The ADC did not have the H369HS in sight and there was no vertical, lateral of longitudinal separation being applied.

The ADC issued a landing clearance to the pilot of the SD3-60 when the aircraft was south of the aerodrome. The ADC requested the pilot of the H369HS to report position and was advised that the helicopter was approaching the mouth of a creek which was 2.5 NM south-east of the aerodrome; on final for runway 32. The ADC instructed the pilot of the SD3-6, which was on left base, to make a right orbit. The orbit was intended to maintain separation between the SD3-60 and the H369HS. The pilot of the helicopter reported sighting the SD3-60 and was then instructed by the ADC to pass behind that aircraft. There was a breakdown of separation.

ANALYSIS

The ADC became distracted by the request for additional circuits and the no radio exercise for the PA28. He did not sight the helicopter and did not apply visual separation prior to approving the pilot of the SD3-60 to make a visual approach. The ADC could have used the tower co-ordinator to assist in sighting the helicopter prior to the SD3-60 being cleared to make a visual approach. This may have enabled the helicopter to be sighted in sufficient time for the application of either a visual or an alternative separation procedure.

SIGNIFICANT FACTORS

1. The ADC did not sight the H369HS.

2. The ADC did not request assistance to sight the H369HS.

3. The ADC did not apply a separation standard between the SD3-60 and the H369HS.

The pilot reported that on arrival over the landing area he noted there was a south-easterly wind of about 30 knots. He therefore planned to land into the south-east. He conducted a low pass in the planned landing direction at about 50 ft above the ground and observed there was no turbulence. He then made an approach at 80 knots with 20 deg of flap selected. At approximately 40 to 50 ft above the ground the aircraft encountered a high rate of sink Despite the application of full power and backstick the aircraft hit the ground heavily, dislodging the nose gear. It then slid for about 25 metres before turning over and coming to rest inverted.

The pilot said there was a hill to the south of the landing area, and he was aware this hill could cause turbulence when there was a southerly wind blowing. However, on this occasion he encountered very strong shear which he now believes is unpredictable in that he did not encounter it when he made the low pass prior to his landing approach.

During a training flight the pilot noticed that the landing gear "in transit" light remained illuminated after the gear retracted. Recycling the gear rectified the problem.

When the aircraft returned to the circuit area air traffic control (ATC) queried the pilot as to whether he had a safe indication for the landing gear. The gear was in the retracted position, but the pilot extended it, and a safe indication was obtained.

Three circuits and landings were then carried out before the aircraft was taxied back to the parking area. The pilot reported having difficulty in maintaining directional control, and thought the aircraft may have suffered a flat tyre.

An inspection of the aircraft revealed that the nut had unscrewed from the right main landing gear leg rear attachment bolt, allowing the bolt to withdraw and the gear to droop. This prevented the gear from retracting completely, which ATC had mistaken for a partially extended landing gear.

The additional load placed on the forward attachment bolt during gear operation and landings caused severe distortion and bending of the front spar web adjacent to the gear attachment fitting.

The reason for the nut to unscrew off the bolt could not be determined. The pilot failed to notice the missing nut during the pre-flight inspection.

The pilot was conducting solo circuit training and had completed 8 normal landings with no problems. He reported that the weather conditions were fine with no noticeable crosswind component on the runway in use.

He stated that the last approach was normal, but as he flared for landing the aircraft continued descending and touched down hard and bounced. He attempted to raise the nose to cushion the landing, but there was no response from his elevator inputs. The aircraft again contacted the ground heavily in a nose low attitude, shearing the nose landing gear attachment frame from the forward fuselage bulkhead, pushing it upwards and shattering the windscreen. Both propellers suffered ground strikes bending the blades.

The pilot may have reduced engine power too soon prior to the landing flare, resulting in a reduction of propeller slipstream over the tailplane, with a subsequent loss of some elevator authority.

The balloon was conducting a passenger carrying flight from Picton to the Campbelltown area, with an intermediate landing at Menangle. The forecast wind was a 14-knot southerly at 1,000 ft. However, during the first flight the pilot determined the prevailing wind to be calm below 500 ft. During the second stage, the pilot noted that the wind above 500 ft was backing to the west although its speed was not increasing.

With the balloon now heading towards the Holsworthy Forest, the pilot decided to land early and made an approach to a large paddock beyond a small dam next to a road. The pilot subsequently reported that, late in the approach, the wind speed suddenly increased to about 15 kt accompanied by a strong wind shear. He also encountered strong downdraughts as the balloon crossed a tree line alongside the road. He attempted to overshoot from the approach, but the balloon continued to descend.

The balloon touched down firmly but was carried into the dam by the wind. A nearby powerline held up the balloon envelope as the basket drifted across the dam and grounded on the bank. Neither the pilot or any of the passengers were hurt.

The pilot said that he loaded the aircraft at the north-western end of the 1,000 m strip oriented south-east/north-west. The strip was partially sheltered from the wind by bordering trees. There was a strong, gusting northerly wind blowing. The pilot assessed that there would be a slight tailwind component for a take-off on runway 12 but believed it was acceptable.

During the initial climb after take-off, the aircraft was subjected to a strong gust of wind which caused it to yaw to the left. It was then struck by another gust. The left wing dropped, and the pilot was unable to regain control before the wing struck the ground. The aircraft cartwheeled onto the right wing and came to rest upright.

An assessment of the weather conditions existing at the time of the accident was made by the Bureau of Meteorology. The wind was from the north at 12 kts with gusts to 19 kts. The temperature was 30 degrees C which meant that the density altitude for the take-off was about 1,800 ft.

The presence of trees which sheltered the strip indicated that turbulence and/or windshear was likely above the tree line. This could have caused the aircraft to lose performance as it climbed above the level of the trees to the extent that it stalled. The height the aircraft was at was probably insufficient to allow recovery before contact with the ground.

As the aircraft took off from Sydney bound for Melbourne, sparks were seen in the area of the right main landing gear. Rescue and firefighting services were placed on local alert for the aircraft's arrival at Melbourne. As the aircraft slowed at the end of the landing roll, the attending firemen reported smoke coming from the right main landing gear. The aircraft was halted while the fire crews extinguished a small fire and reported that one tyre was deflated. The engines were shut down and the aircraft towed to the terminal. Fire crews remained in attendance and reported that the tyre began to smoulder as the aircraft was approaching the parking bay. The passengers were deplaned normally. Forty minutes after landing the situation was reported under control.

Initial maintenance investigation found that the inner bearing on the number 8 wheel had seized. Extensive secondary damage had been sustained by the tyre, wheel and bogie unit. Further examination disclosed that the number 4-wheel inner bearing, fitted to the left bogie, showed similar damage.

The damaged components were returned to the operator for investigation under control of the Austrian airworthiness authority (AustroControl). The results of the metallurgical assessment of the bearing damage have not been received to date.

The manufacturer advised that there has been a small number of failed bearings on B767 aircraft. A service letter (767-SL-32-070) has been issued to introduce a grease seal and grease dam to resolve the problem of abnormal wear, overheating and damage to wheel bearings caused by inadequate bearing lubrication. To date the operator has not advised whether the bearings that failed were modified to the service letter standard, although AustroControl advised that the pattern of wear observed on the failed bearings, plus the presence of a quantity of grease residue would tend to preclude lack of lubrication as a causal factor.

The service letter also advised that the existing grease type, Aeroshell 5, was replaced with Aeroshell 22 or Mobil 28. The service letter advised that Aeroshell 5 could still be used on B767 aircraft, but not on B747 aircraft. To date the operator has not advised which grease was in use on the failed bearings.

There is no information from the operator on these aspects, but because the service bulletin was issued on 29 August 1996, only 8 weeks prior to this incident, it is considered unlikely that its requirements had been incorporated into the failed bearings.

The manufacturer also prescribed a critical wheel bearing torquing procedure that required double torquing of the bearings. AustroControl advised that the double torquing procedure is well known and observed within Lauda Air, and that a recent audit of the brake and wheel shop in accordance with JAR145 found that all pertinent procedures were being followed. The factors surrounding the failure of the 2 bearings have not been determined.

FACTUAL INFORMATION

History of the flight

A Boeing 747 (B747) was en route from Sydney to Osaka, tracking via Narrabri, Hamilton Island and Port Moresby. The crew had originally planned at flight level (FL) 310 with the intention of climbing to FL350 prior to entering Papua New Guinea airspace. After departing Sydney, the crew calculated that the aircraft could immediately climb to FL330 and requested a change to that level. This was a non-standard level for the planned track. As the B747 flight was to be conducted under radar control while in Australian airspace, the controller granted approval for the crew to operate at FL330. Subsequently, the change to a non-standard level, for the track being flown, was co-ordinated with all other controllers responsible for the Australian airspace through which the B747 would pass.

A Boeing 737 (B737) had departed from Cairns on a flight to Brisbane and was tracking direct to Mackay at FL330. This aircraft was operating at a standard level for the intended track.

The B747 entered the airspace under the jurisdiction of the Brisbane Sector 7S radar controller as the aircraft passed 75 NM to the west of Rockhampton. To ensure adequate coverage of the sector, the controller was using the 180-NM scale on the radar display. Within the sector, the track of the B747 was to cross five other routes which either converged or intersected. Aircraft using these other routes could operate at standard and non-standard levels relative to the track of the B747.

Sector 7 had two radar positions - Sector 7S and Sector 7V. Additionally, there was a planner position located between these radar positions. The planner controller was assisting the Sector 7V controller and not the Sector 7S controller.

The actual time the B737 entered the Sector 7S area, south of Townsville, was not determined. However, based on groundspeed calculations made by the investigation team, the B737 was estimated to have entered the sector approximately 5 - 10 minutes after the B747. The controller was busy at the time and satisfied with the overall traffic situation but did not appreciate the possibility of the two aircraft coming into conflict. The B747 was at FL330 as it approached Hamilton Island from the south. The tracks for the two aircraft crossed 33 NM south-south-east of Hamilton Island. As the aircraft approached the crossing point, the radar controller was required to coordinate and separate a number of departures from Mackay and Hamilton Island airports.

The horizontal separation between the B747 and the B737 aircraft had reduced to 5.5 NM before the controller observed the proximity of the two aircraft on the radar display. The controller instructed the crew of the B737 to turn right to pass behind the B747.

The crew of the B737 sighted the B747 as the controller issued the avoidance instructions. The crew had previously observed the B747 in the distance, but had not perceived it as an aircraft. They were about to request clarification from the controller about a possible aircraft approaching them, when they were advised to turn right. The B737 crew complied with and acknowledged the instruction. The controller subsequently issued traffic information on the B747 to the crew of the B737.

The crew of the B747 received a Traffic Advisory (TA) warning on the aircraft's Traffic Alert and Collision Avoidance System (TCAS), which indicated traffic at the 10-o'clock position at 4 NM. The crew looked for the traffic but did not sight the B737.

The two aircraft passed with a horizontal separation of approximately 2.4 NM and at the same level. The minimum separation standard required was 5 NM horizontally or 2,000 ft vertically. There was a breakdown of separation.

Aircraft cruising levels

Normally, aircraft are approved to operate at flight levels in accordance with the instrument flight rules (IFR) cruising level table. The table used provided standard vertical separation between aircraft which were flying on easterly (example levels are FL330, FL370 and FL410) and westerly (example levels are FL310, FL350 and FL390) magnetic tracks. Approval to operate at other than standard levels could be granted by an air traffic controller when traffic or other circumstances required a change in level assignment. In order to assign a non-standard level, controllers are required to consider the implications on workload and coordination, and the effect on other aircraft which were operating at standard levels. These aspects were considered prior to the B747 crew receiving approval to operate at a non-standard level.

Sector 7S

The Sector 7S controller was responsible for the provision of en-route control services for transiting aircraft as well as arrival and departure control services for aircraft inbound/outbound from airports within the sector which covers a large portion of the Queensland central coast. Additionally, the controller was responsible for the provision of a radar advisory service (RAS) and a search-and-rescue (SAR) alerting service for aircraft operating in non-controlled airspace within radar coverage in the eastern portion of the sector.

Controllers were aware of the potential problems with the crossing routes on this sector, and similar situations were practised in the simulator and encountered on a regular basis when operating the position.

Sector 7S controller

The controller had two and a half years experience in air traffic control and had recently passed a proficiency assessment. He had worked the same shift period the previous day and was adequately rested.

He had been on duty in the position for approximately 50 minutes prior to the occurrence, during which there had been a steady increase in traffic. He appreciated the level and complexity of the traffic situation at the time and felt comfortable with his control performance. He had not considered requesting assistance from the planner controller.

Sector 7 planner

The Sector 7 roster had recently been amended to facilitate staffing of the planner position during nominated times. These times covered anticipated busy periods when the planner controller would be of assistance to the two radar positions. The planner controller conducted coordination with other air traffic service (ATS) agencies on behalf of both radar controllers. This enabled the radar controllers to concentrate on the separation and management of traffic within their respective areas of responsibility.

When all three positions were staffed, the planner and radar controllers shared a communication line to a number of approach control centres serving airports within the two radar sectors. When this line was being used by any of the Sector 7 positions for coordination, it could not be used by either of the other two positions. Consequently, coordination to and from other ATS agencies was often delayed until the line became available. Also, the planner position had facilities to enable a controller to monitor the air-ground-air program of the two radar positions. The planner controller could monitor both radar positions concurrently.

Subject to workload, the planner controller could observe the performance of the radar controllers and provide some assistance to separate traffic if required. The planner controller was required to manage the assistance provided to any one radar controller to ensure it was not to the detriment of the other radar controller.

Traffic situation

The tracks and levels of the departing aircraft from the Mackay and Hamilton Island airports required, the Sector 7S controller to employ step-climb procedures. Subsequently, he spent some time ensuring separation between a number of aircraft in the area to the immediate south of Mackay. At the same time, he was monitoring another radio frequency expecting a transmission from the pilot of an aircraft which was due to arrive at Shute Harbour. The controller also had a number of other aircraft throughout the sector operating on both the control and RAS frequencies.

ANALYSIS

Flight routes and cruising levels

Generally, the operation of aircraft at levels in accordance with the IFR cruising level table would have provided the standard vertical separation of 2,000 ft between the two aircraft. However, because the route of the B747 intersected a number of north and southbound air routes, it would have conflicted with one of the routes, no matter what level was maintained. Action to separate aircraft on the various crossing routes was required on a regular basis. This required one aircraft to operate at a non-standard level or to be radar vectored until the situation was resolved.

As the B747 was to cross a number of routes which may have other aircraft at the same level, the safety net provided by the use of the cruising table levels was not available.

Sector 7 radar controller

The B747 and the B737 entered the sector at its southern and northern extremities respectively. The distance between these entry points (approximately 160 NM on the radar display), possibly made it difficult for the controller to appreciate the future potential for conflict between the aircraft. Also, the controller was dealing with a number of aircraft in the area south of Mackay and was distracted from regularly scanning the full display. These aspects combined to create a situation where the controller did not develop an awareness of the potential conflict.

The provision of assistance from the planner controller may have reduced some of the workload and enabled the radar controller to widen his scan of the sector. This may have enabled him to recognise the pending conflict between the B747 and the B737. Alternatively, the planner controller may have recognised the potential conflict and alerted the radar controller prior to the horizontal separation reducing to the minimum. However, because the planner controller was assisting the Sector 7V controller, she was unable to assist the Sector 7S controller or monitor his display. The radar controller thought he was coping adequately with the situation, and consequently did not request any assistance.

After resolving the situation near Mackay, the radar controller turned his attention to other areas of the sector. He quickly realised the situation and provided instructions to the crew of the B737 to avoid the B747. Generally, when there is a breakdown in separation, the provision of traffic information to the crews of the aircraft involved has priority. However, in this incident the priority was to have the B737 crew turn their aircraft away from the B747. Under the circumstances, the radar controller's momentary pause to receive an acknowledgment from the crew of the B737 prior to issuing traffic information was judicious. By this action, he was able to assure himself that the B737 crew had received the essential avoidance instructions before transmitting the traffic information.

The large scale and the variety of aviation activities occurring within the Sector 7S area made management of the sector difficult at times. The sector controller was required to resolve a number of conflicts within the sector concurrently. This resulted in the controller's attention being focused on one particular area of the display while separation action was being implemented. Consequently, other areas of the display did not receive adequate monitoring.

Human factors considerations

The controller was required to provide en-route control services to high-level transiting aircraft, an arrival/departure control service to aircraft operating to and from the underlying airports, and a RAS and SAR alerting service. The first two services are similar in implementation but generally cover different height bands. However, the provision of a RAS and a SAR alerting service represents significantly different types of tasks in cognitive terms compared to the other services. A study of United States Federal Aviation Authority air route traffic control centres (Bruce and colleagues,1993) indicated that a controller's task load is not solely related to increasing traffic, but is also very much conditional upon the degree of change in complexity of the overall traffic situation. Such was the case in the leadup to this incident.

A controller needs to be able to recognise the change in task complexity as well as an increase in traffic activity to manage the overall task better. Training for controllers to develop a specific awareness of when they are approaching task saturation would be a defence for future incidents. Also, resources should be readily available to provide assistance when controllers recognise that they are approaching task saturation.

For most of the time the controller may have been capable of readily providing all the required services concurrently. However, the differences in the types of the services provided and the spread of the traffic in the vertical and lateral planes across the sector, resulted in the controller requiring a complex management plan to be able to adequately manage the task. Consequently, because of this complexity, the total task load increased to the point where the controller may have become task-saturated.

Sector 7 planner

During busy periods, the planner controller could only provide limited assistance to the two radar controllers. This was mainly due to the limitations of the single available communications line to some of the other ATS agencies. The mode of operation of the communication line reduced the flexibility for the planner and the radar controllers to conduct timely coordination. Often when the line was available, the controllers were busy communicating with aircraft or conducting coordination with other Brisbane ATS positions. Provision of separate communication lines to the two radar and the planner positions would enable all three to conduct coordination concurrently. This would provide more options to reduce the task load.

The planner controller was the only immediate "safety valve" available to either radar controller should the latter approach an overload situation. Modification of the planner position to enable another controller to assist at the position may be warranted. This would ensure that assistance was readily available to either radar position which in turn would limit the possibility of future controller task saturation.

SIGNIFICANT FACTORS

1. The planned route of the B747 crossed a number of other routes which could have other traffic at the same level.

2. The controller was responsible for the provision of different services within a sector in which a number of areas required close monitoring concurrently.

3. The controller's attention was focused on separating traffic located immediately to the south of Mackay to the detriment of maintaining a regular scan throughout the total area of his responsibility.

4. The controller believed he was coping with the situation and did not consider requesting assistance from the planner controller.

5. The planner controller was assisting the Sector 7V controller and was unable to assist the Sector 7S controller.

SAFETY ACTION

This investigation identified specific safety deficiencies associated with the provision of ATC services in Brisbane Sector 7. The deficiencies were related to the complexity and workload of the Sector 7 radar controller, to the provision of only one planner controller to assist the two radar controllers and the limited communication facilities for the planner to assist the radar controllers.

The complexity and workload issues were compounded by the Sector 7 radar controller being responsible for a RAS as well as an en-route ATC service. This issue has been addressed by IR960009 which was issued on the 14 August 1997. The Interim Recommendation stated:

"The Bureau of Air Safety Investigation recommends that Airservices Australia re-assess the safety implications of providing a radar advisory service in conjunction with a radar control service."

On 15 September 1997, Airservices Australia responded to the draft occurrence report and advised that they had reviewed the implications of providing a radar advisory service in conjunction with a radar control service and were satisfied with the service currently provided in this sector.

Airservices Australia also advised that team leaders were rostered from 0600 to 2000 hours daily who were able to monitor the workload and complexity of each position and were able to take appropriate action to maintain the integrity of the positions.

The provision of additional facilities was addressed by Safety Advisory Notice (SAN) 970130 which was issued on the 17 September 1997 to Airservices Australia and the Royal Australian Air Force. The SAN was related to this and a number of other occurrences and identified the following safety deficiency:

"Aircraft movement coordination between Brisbane Sector 7 and Townsville ATC operator positions is constrained at times by the single inter-communication line."

Airservices Australia have informally advised that they intend to modify the two Sector 7 radar consoles to provide independent satellite communication facilities in late 1997 in conjunction with other scheduled engineering modifications.

The overall aspects of the ATC task methodology and human performance will be examined in a detailed study by the Bureau of Air Safety Investigation.