Loss of separation

1. The similarity of the departure instruction to the noise abatement procedure provided the potential for a human error to occur.
2. The crew of TAF did not comply with the departure instruction issued by air traffic control because the crew confused the instruction with the noise abatement procedure.
3. The enroute controller did not issue TAF with a procedural clearance that would have provided separation assurance with CZV.

The investigation determined that the crew of TAF did not comply with their departure instruction. Separation would not have been infringed if TAF had continued on a heading of 090 M as directed, instead of intercepting the 128 radial. The similarity of the noise abatement procedure to the departure instruction may have diminished the importance of the requirement for the crew to maintain a heading of 090 M after departure.

Although the action of the crew resulted in an infringement of the radar separation standard, the enroute controller did not issue a clearance that assured separation between TAF and CZV. Separation assurance was required in accordance with MATS 4-1-1, paragraph 6. The controller used radar procedures to separate the two aircraft, where non-radar separation should have been applied in accordance with MATS 4-1-1 paragraph 11 because communications and equipment did not allow the application of radar separation.

The enroute controller was not in direct communication with the crews of both aircraft. Moreover, portions of Maroochydore airspace were below radar coverage including airspace below 1,500 ft. Although LOA 98/054 allowed the enroute controller to "assume radar identification of departing aircraft within four minutes of the next call", published procedures did not permit the enroute controller to apply radar separation between radar identified aircraft and aircraft departing Maroochydore that were not identified.

The crew of VH-TAF, a Boeing 737, was taxiing for runway 18 at Maroochydore (MC) for a departure to Sydney. The route was flight planned, and subsequently cleared, via waypoint TRIKI (128 MC at 22 NM) on air route W196 at flight level (FL) 350. The crew had pre-briefed the noise abatement procedure that was detailed in AIP Australia as follows:

"2.2 Departing Runway 18 - MC to TRIKI: UNLESS OTHERWISE DIRECTED BY ATC. (a) Jet aircraft shall at 500FT, turn left heading 090 degrees M to intercept track. If not able to intercept track from this heading, then maintain heading 090 degrees until 4DME before turning to intercept track".

Another Boeing 737, VH-CZV, was inbound to Maroochydore from Sydney and tracking via TRIKI on W196. The Maroochydore aerodrome controller had received prior coordination on this aircraft, which was on descent to 5,000 ft. Although the BURNET enroute controller was responsible for CZV, the crew was not in two-way communication with the enroute controller. The enroute controller had agreed to allow the crew to transfer frequency direct from Brisbane Approach to Maroochydore Tower. The standard frequency transfer point in accordance with Letter of Agreement (LOA) ND 98/054 was TRIKI. CZV was roughly 32 NM from Maroochydore when the aerodrome controller contacted the enroute controller to coordinate a departure clearance for TAF.

Maroochydore was a non-radar tower and the aerodrome controller was responsible for providing procedural separation to aircraft below 4,500 ft. The enroute controller was responsible for the surrounding airspace overlying and to the south of Maroochydore as defined in the AIP Designated Airspace Handbook. The enroute controller was able to use radar or procedural means to provide air traffic control separation. Radar coverage in the Maroochydore area was generally available above 1,500 ft.

At 1236, the enroute controller advised the aerodrome controller to issue TAF a heading of 090 degrees M and to maintain FL150, a level that would provide vertical separation from a third aircraft overflying at FL160. The aerodrome controller recognised a potential conflict between TAF and CZV, and confirmed with enroute that the enroute controller would accept the responsibility for conflict resolution. The enroute controller accepted the responsibility for providing separation.

The aerodrome controller then issued the following departure instruction to the crew of TAF: "TAF, restriction on departure is to turn left and take up a heading of 090 magnetic, maintain FL 150, clear for take-off, make left turn". The crew correctly read back the instruction. The crew later reported that they interpreted the restriction on departure to be the altitude restriction and believed the heading instruction applied to the noise abatement procedure. At 1238, the crew was instructed by the aerodrome controller to contact the enroute controller to make a departure report.

CZV meanwhile, was approaching TRIKI and the crew reported to the Maroochydore aerodrome controller on descent to 5,000 ft. The aerodrome controller directed the crew to continue tracking direct to Maroochydore and to maintain 5,000 ft. Maroochydore tower then contacted the enroute controller seeking an unrestricted descent clearance for CZV, which was granted but not issued to the crew.

At the same time, TAF was turning right to intercept the 128 radial at 4 NM from Maroochydore and was passing 4,400 ft on climb. The crew contacted the enroute controller at 1239 and reported established on the 128 radial, climbing to FL 150. TAF and CZV were then on reciprocal tracks and closing. In an unsuccessful attempt to preserve the required separation, the crew of TAF was told to turn left onto a heading of 090 M and maintain 5,000 ft. The aerodrome controller who was in two-way communication with the enroute controller over the hotline, found that CZV was passing 5,400 ft on descent. The enroute controller asked the aerodrome controller to direct the crew of CZV to turn right on to 090 M, which was actioned. The enroute controller issued traffic information to the crew of TAF about the relative position of CZV and requested the crew to climb to FL 150 and maintain their best rate of climb. Traffic information was not issued to the crew of CZV about TAF.

At 1240, the required vertical separation of 1,000 ft between the two aircraft reduced to 400 ft when the minimum radar separation of 5 NM was infringed. The application of radar separation and the requirements for issuing clearances were detailed in chapters 4, 9 and 12 of the Manual of Air Traffic Services (MATS).

The enroute controller received a short-term conflict alert (STCA) on the radar display. However, neither of the flight crews reported receiving a traffic alerting and collision avoidance system (TCAS) warning from their respective aircraft's display.

1. The supervisor of the trainee, the aerodrome controller, had a lapse in concentration at a critical stage of the departure sequence, having been operating at the position in excess of 3 hours.
2. The trainee was unfamiliar with terminal area operations/procedures and was unaware of the increase in complexity during runway transition.
3. Tower training publications' critical safety section and other references did not include potential standard instrument departure or arrival areas of conflict.

The concentration level of both the aerodrome controller and the trainee had probably lapsed due to the duration of the training session. Additionally, the aerodrome controller may have been less vigilant in his monitoring of the trainee because of the reduction in aircraft numbers and his perception of the trainee's level of competence in the aerodrome control position.

Provision of information regarding the increase in complexity of traffic management during runway transition and regarding the critical safety aspects of standard instrument departures and arrivals would have provided the trainee with some knowledge to compensate for areas where he lacked experience. On-the-job training does not necessarily expose trainees to all critical elements nor provide sufficient time for them to develop their own traffic management models. Information on such critical areas is required to ensure that safety aspects are understood and considered, particularly during on-the-job training.

A greater awareness of terminal area operations and procedures may have provided the trainee with sufficient knowledge to appreciate that the departing aircraft were not going to be separated by the departure controller. In turn, this may have alerted the trainee to the fact that he was required to provide a separation solution or alternatively coordinate one with the departure controller.

A Beech 1900 (Beech) departed from runway 27 and was tracking via a STRATHBOGIE 8 standard instrument departure, on climb to 5,000 ft. Shortly after, a Boeing 767 (B767) departed from runway 34 on a DOSEL 8 standard instrument departure, on climb to 5,000 ft. The standard instrument departure tracks intersected approximately 2 NM north of the aerodrome. The departure north radar controller (departure controller) observed that the separation between the two aircraft was not going to be maintained and queried the aerodrome controller in relation to the separation procedure being applied. The aerodrome control position was being operated by a trainee under the supervision of a rated controller. The trainee advised that he thought the departure controller was providing separation. The departure controller instructed the Beech crew to turn left onto a heading of 310 degrees and the B767 crew to turn right onto a heading of 030 degrees. The two aircraft passed at the same altitude, with 1 NM lateral separation. The minimum separation required was either 3 NM laterally or 1,000 ft vertically. There was an infringement of separation standards. The B767 was fitted with a traffic alert and collision avoidance system and the crew received a traffic advisory just prior to the issue of the avoiding instructions from the departure controller.

The trainee and the aerodrome controller had been at the console for about 3 hours and the number of aircraft being managed had reduced slightly following a busy period. The runway mode was being changed from land and hold short operation (LAHSO) using runways 34 and 27, to dedicated departure runway arrangement using those runways. In LAHSO all aircraft depart from runway 27, while in the dedicated departure mode, departures to the south and west use runway 27 and those to the north and east use runway 34. In that configuration, the intersection of aircraft departure tracks is almost eliminated. It was recognised by experienced controllers that the transition period between runway modes required an increased awareness and careful consideration of the traffic management plan to ensure that potential conflicts for departing aircraft were minimised. The trainee was aware of potential problems for each mode but had not been alerted to the potential problems that may occur during a mode transition period. On taxi, the crew of the B767 requested and were approved to use runway 34. The trainee subsequently amended the departure order of the B767 and the Beech to eliminate a requirement to apply a 2-minute wake turbulence separation standard to the Beech aircraft.

The trainee did not appreciate that the aircraft's departure tracks converged north of the aerodrome. The aerodrome controller reported that as the trainee was soon to be checked for a rating his monitoring of the trainee's control was less vigilant than normal. Local instructions stated that the aerodrome controller was responsible for the separation of departing aircraft within 5 NM of the aerodrome and that for departures from different runways, the aerodrome control and departure control positions were jointly responsible for coordinating a separation solution. The departure controller's preference was to depart the B767 before the Beech but he agreed to the trainee's request to change the departure order. Normal practice in this situation was that the aerodrome control position, having initiated the change, was responsible for the separation unless otherwise coordinated. The controllers neither coordinated a separation solution nor clearly established who was responsible for ensuring separation. Both assumed that the other was going to separate the two aircraft. The aerodrome controller was not aware of the developing situation and was only alerted to the infringement following the departure controller's query to the trainee.

The trainee had undergone tower control ab initio training, which included approximately 1 day of training on the vectoring of aircraft. The course did not provide any terminal area (TMA) control training. Controllers undergoing training for tower positions were able to undertake TMA familiarisation. But the tower training syllabus did not require trainees to develop an understanding of, or to be familiar with, TMA operations. TMA familiarisation by tower trainees was conducted on an ad-hoc basis and there was no procedure to assess a trainee's level of understanding of TMA operations.

Melbourne Tower training publications did not provide advice of a preferred duration for training sessions that would maximise the training benefit. The publications did list aspects of critical safety that needed to be covered during training for both trainees or supervising controllers. But the list did not include areas of known conflict for standard instrument departures or standard arrival routes. The publications did not include advice of the potential inherent complexity of runway mode transition.

This scenario only occurred prior to first light, which occurred after 0600 from approximately the end of March until early September. The problem with the SIDs converging during daylight hours was normally overcome by the cancelling of the SID and the issuing of a radar departure instruction while the aircraft is on the ground prior to departing RWY 03, thus providing separation assurance.

Prior to first light, separation assurance could have been achieved by DEP applying vertical separation between aircraft departing these runways, or by exercising control over aircraft release times to ensure that a radar standard would exist between them.

The DEP controller did not ensure separation between the B737 and the preceding BAe146 by permitting the trainee to issue an instruction of "unrestricted" to the ADC.

The crew of the BAe146 provided a prompt to the controller by requesting a clearance direct to Ballidu shortly before commencing a turn off runway heading. However, the crew were advised by the DEP trainee that the request was understood and that he "will advise".

A second prompt occurred to the training officer when the crew of the B737 called on the DEP frequency reporting that they were leaving 1,800ft. It appears that this prompt alerted the training officer to the developing situation, as he then instructed the trainee to turn the BAe146.

The training officer at this point still did not intervene directly, but instead indicated to the trainee that he needed to give the crew of the BAe146 a heading in order to resolve the situation more quickly. The provision of traffic information or use of "immediate" may have been appropriate under the circumstances.

That the training officer did not intervene until after the incident had largely been resolved was indicative of inattention to the operational situation and a lack of understanding of the criticality of the situation. The training officer displayed a lack of control over the trainee at a time when the issue of precise instructions and a need to alert the crew of the situation was critical. Instructions issued to the trainee were in general terms rather than the specific terminology that the situation demanded.

The task of the ADC in this instance was to manage the runway departures in accordance with instructions from the DEP controller. The ADC and ADC trainee, both understood that the primary responsibility for separation for aircraft on the 03 and RWY 06 PEPPA3 SID, lay with the DEP controller.

The ADC advised that he continued to visually monitor both aircraft via their departure tracks, even though responsibility for separation lay with the DEP controller. He also advised that at no stage did he offer to take, or accept any, responsibility from the DEP controller for separation of the aircraft.

When asked by the DEP controller whether he could monitor, the response of "Yes, I'm monitoring" was intended to indicate that he had the aircraft in sight and that, at that time, by visual observation their tracks had begun to diverge. He was thus indicating that he was able to monitor that the aircraft tracks were not likely to come together, but was not indicating any acceptance of responsibility for separation.

The training configuration utilised by the ADC and DEP was such that the trainers were unable to directly override their trainees. Use of available override systems would have made it easier for the trainers to readily communicate with the crews.

1. The controller did not withhold the take-off clearance for CZO when there was no reasonable assurance that separation would exist between the aircraft when CZO commenced the take-off roll.
2. The lack of visibility of the A4S taxiway at night from the control tower required that aerodrome controllers instruct pilots to confirm when clear of the runway.

The difficulty in accurately locating aircraft on the high-speed taxiway A4S at night from the control tower is evidenced by the controller twice assuming that OGG had departed the runway and entered the taxiway. The awareness among local controllers of this difficulty should have encouraged the use of procedures to ensure the maintenance of required standards. However, the controller only recognised that his visual assessment was incorrect when advised by the crew of OGG. It is likely that the controller subsequently assessed that as CZO had not lined up, it was not necessary to cancel the clearance, but the two aircraft should simply be monitored.

The procedures relating to take-off clearances permitted a degree of discretion by the controller. However, that discretion was conditional upon a valid determination of collision risk and the maintenance of separation standards. Consequently, MATS 6-2-3 paragraph 32 was not appropriate to the issuing of the take-off clearance, as the controller had not positively identified the relative positions of the aircraft.

1. The runway 15 SWIFT standard instrument departure allowed the possibility of a following aircraft turning inside a preceding aircraft.
2. The performance of the B737-400 series aircraft was superior to that of the B737-300 series aircraft.
3. The use of minimum departure separation standards was inappropriate.
4. The decision of the approach/departures controller to concentrate on data entry tasks during the departure sequence was not sound practice.

Controllers at Cairns considered that the aircraft were "like types" for the purposes of departure standards and neither the aerodrome controller nor the approach/departures controller considered increasing the separation requirements specified in Local Instructions. However, Boeing 737-400 series aircraft are known to normally out-perform the Boeing 737-300 series.

Separation standards used for consecutive departures allowed controllers to increase the minimum distance at their discretion. On this occasion it was not increased and, although the runway departure standard was not breached, the decision did not allow for any unexpected manoeuvre by either aircraft. Both crews acted in accordance with the standard instrument departure and the closure was the effect of aircraft performance.

The approach/departures controller elected to input data to the air traffic computer during the departure sequence. Although these actions were necessary, they were labour intensive and diverted his attention from the air situation display. They were not urgent and the decision resulted in the controller not having his full attention on the relative positions of the aircraft when they first appeared on the display.

The design of the SWIFT 2 standard instrument departure did not guarantee separation assurance. Whenever the second aircraft reached 4,000 ft prior to the first aircraft (whatever the reason) a reduction in horizontal separation was likely.

As the flow controller was planning the arrival sequence, he needed to take account of the type of approach available to each aircraft. Even though visual approaches from the south-west had been in use during the morning, tower staff had advised the terminal area controllers that they were no longer available. However, the flow and approach controllers elected to continue with the plan that had the C310 as number one in the sequence and hoped that the pilot would be able to complete a visual approach. The conversations between tower and terminal area controllers in regard to the status of aircraft approaches took place between various control positions, each in partial isolation. Consequently, there was no coordinated response between tower and terminal area staff, and some confusion arose as to the status of such approaches.

As the C310 had been the first aircraft in a busy traffic sequence, the C310 needed to be repositioned when the pilot was unable to make a visual approach. When this re-sequencing occurred, pressure was on the system to try to make up some time to allow for the re-introduction of the C310. In order to minimise the delay, the approach controller had attempted to reduce the distance between the C310 and the AC50 by use of techniques that were not appropriate for the prevailing weather conditions.