Loss of separation

At 1526 Central Standard Time on 4 September 2012, a Bombardier DHC-8-202, registered VH-ZZJ (ZZJ), was cleared to take-off on runway 29 at Darwin airport, Northern Territory. At 1528 and before ZZJ had commenced take-off, an Aerospatiale AS.332L1, registered VH-LOJ (LOJ), called ready on runway 36. The Aerodrome Controller (ADC) cleared LOJ for take-off and instructed the pilot to maintain upwind.

The pilot of ZZJ heard the transmission and, believing that there would have been a risk of a collision if he commenced take-off, delayed his departure until after LOJ had crossed the intersection of the two runways. The action of the pilot of ZZJ to delay take-off reduced the likelihood of a more serious incident.

Air traffic services at Darwin are provided by the Department of Defence (Defence). The Defence investigation found that an unsafe take-off clearance had been issued to LOJ, resulting in a runway separation breakdown. Further, the ADC failed to monitor both ZZJ and LOJ during a critical phase of operation – take-off. The actions of the pilot of ZZJ, by delaying take-off, reduced the severity of the incident.

The pilot of ZZJ did not query the ADC’s instructions at the time and did not report the incident until two days later. The pilot of LOJ was not aware of the incident and the ADC did not recall the incident.

Aviation Short Investigation Bulletin – Issue 17

On 9 August 2012 at 1235 Eastern Standard Time, a breakdown of separation occurred between a Raytheon Beech 200, registered VH-FDD (FDD) operating under the callsign FLYDOC 423, and a Raytheon Beech 350, operating under the callsign DINGO 008, near Cooktown, Queensland. An air traffic controller had issued clearance for FDD, tracking north, to climb to flight level (FL) 300. During the climb, the air traffic control system’s Short Term Conflict Alert (STCA) activated, when the aircraft were 10.1 NM horizontally and 200 ft vertically apart. The STCA alerted the controller to a pending breakdown of separation with the southbound aircraft, DINGO 008, maintaining FL270. The controller issued both flight crews with safety alerts and instructions that limited the severity of the separation breakdown. The aircraft passed 4.8 NM horizontally and 100 ft vertically of each other.

The controller had not identified the conflict before the STCA activation, but their knowledge and application of effective compromised separation recovery techniques was integral to resolving the issue, following their recognition of the situation.

This incident highlights the importance for controllers to effectively balance their professional desire to promptly facilitate pilot requests with the overriding requirement to provide a safe and efficient air traffic control service.

The prompt and effective controller reaction to re-establish the appropriate separation standard highlights the benefit of and importance of regular compromised separation recovery training as an integral defence.

Aviation Short Investigation Bulletin - Issue 13

On 21 June 2012, a breakdown of separation occurred between a Qantas Airways Boeing Company B737-438, registered VH-TJS (TJS) and a Sharp Aviation Fairchild SA227, registered VH-MYI (MYI) near Melbourne Airport, Victoria. After being radar vectored from the north-east and given an assigned heading of 230˚, TJS was cleared to conduct an instrument approach to runway 27 at Melbourne Airport. However, instead of intercepting the runway 27 localizer, TJS maintained the assigned heading and passed through the localizer by about 1 NM. As a result the separation between TJS and MYI reduced to less than the required standard of 3 NM.

Although the reason for the clearance deviation could not be positively identified, it was likely that a higher than usual workload resulted in the flight crew not arming the appropriate approach mode or observing the aircraft’s closure relative to the runway 27 localizer. The occurrence highlights the need for flight crews to closely monitor the aircraft flight path and be ready to take prompt action in the event of any deviation from the cleared route.

Aviation Short Investigation Bulletin - Issue 13

On 21 June 2012, after a review of the initial Airservices Australia report of the occurrence, the Australian Transport Safety Bureau (ATSB) initiated an investigation into a breakdown of separation (BOS) between two aircraft near Melbourne Airport, Victoria on 20 June 2012.

A preliminary examination of the occurrence showed that the BOS involved an Airbus A320 aircraft (A320) that was inbound to Melbourne Airport from the north and a Bombardier Inc BD-700 aircraft (Bombardier) that had departed from runway 34 at Melbourne and was tracking to the north-west. The ATSB also reviewed information from the involved Melbourne Approach Departures controller, the relevant air traffic control automatic voice recordings, weather information from the Bureau of Meteorology and the respective aircrafts flight tracks on Webtrack recordings (see Airservices WebTrak). 

The preliminary investigation determined that the controller had planned for a lesser rate of closure between the two aircraft than eventuated, with the actual rate of closure being affected by the ambient conditions and a slower acceleration by the Bombardier than anticipated by the controller. No avoiding action was necessary by either crew. The controller became aware of the infringement of the 3NM (5.6km) separation standard when advised by the crew of the A320 that they had received a Traffic Alert and Collision Avoidance System traffic advisory on passing behind the Bombardier. Shortly after, the lateral and vertical distance between the two aircraft increased and a separation standard was re-established.

The preliminary investigation showed that the potential for any systemic issues to have contributed to the occurrence was low and that it was unlikely that any safety issues would be identified through further investigation. On that basis, the ATSB decided to discontinue its investigation, as provided for in Section 21 (2) of the Transport Safety Investigation Act 2003.

On 6 April 2012 a Boeing Company 737-838 aircraft (B737), registered VH-VXI, on a scheduled passenger flight from Sydney, New South Wales to Darwin, Northern Territory was involved in a breakdown of separation with an Airbus Industries A330-243 aircraft (A330), registered B-6073, on a scheduled passenger flight from Melbourne, Victoria to Shanghai, China.

Both aircraft were operating under Instrument Flight Rules and were in airspace controlled by Brisbane Centre utilising radar.

The B737 and A330 were on converging tracks at Flight Level (FL) 360. As the aircraft approached Tindal, Northern Territory at 1342 Central Standard Time, a breakdown of separation occurred when the distance between them reduced to about 3.5 NM, before vertical separation was established. The incident occurred about 16 minutes after a handover between two Air Traffic Controllers.

As a result of a number of incidents involving handovers, Airservices Australia advised the ATSB that they had amended the handover procedure to require supervision and for the relinquishing controller to remain at the console to provide assistance until the accepting controller indicated that assistance was not required.

The incident highlighted the importance of separation assurance and thorough handovers.

Why the ATSB is doing this research

One of the main objectives of air traffic services (ATS) is to prevent the collision of aircraft. Aircraft separation standards are set to ensure that the chance of a mid-air collision is very remote. When they are infringed, there are fewer defences left to guard against a mid-air collision. This ATSB research investigation looks at loss of separation (LOS) incidents in Australian airspace to understand how often they occur and in what contexts, how and why they are occurring, and whether there are any wider implications that the air traffic system is not functioning appropriately.

What the ATSB found

Although there had been an increase in the number of occurrences reported to the ATSB over the 2 years ending in June 2012, there were fewer LOS occurrences during that period than during 2005 to 2008. Traffic levels have generally increased during the same period. A LOS between aircraft under air traffic control jurisdiction happens on average about once every 3 days. In almost 90 per cent of LOS occurrences, there was no or minimal risk of aircraft colliding. On average, however, there are six occurrences per year where an elevated risk of collision exists. There have been no mid-air collisions in Australia between two aircraft under ATS control.

The investigation found that military controlled terminal area airspace in general, and all airspace around Darwin and Williamtown in particular, had a disproportionate rate of LOS (for civilian aircraft). Most of these LOS occurrences were contributed to by air traffic controller actions. This may be a result of the nature of aircraft operations and airspace constraints at some military airports, leading to reduced use and effectiveness of strategic separation defences, thereby placing more responsibility for separating aircraft directly onto the controllers. Furthermore, as military ATS are not subject to safety oversight by the Civil Aviation Safety Authority (CASA), there is no independent assessment and assurance as to the safety of civilian aircraft operations at military airports.

In civil airspace, LOS occurrences attributable to pilot actions are not monitored as a measure of airspace safety nor actively investigated for insight into possible improvements to air traffic service provision. As about half of all LOS incidents are from pilot actions, not all available information is being fully used to assure the safety of civilian airspace.

What's been done as a result

The ATSB has issued recommendations to the Department of Defence to review all processes and controls in place for aircraft separation in military ATS and to CASA to review whether its current level of involvement with military ATS is sufficient to assure the safety of civil aircraft operations. The ATSB also recommends using all available information, including pilot attributable LOS occurrences, to assure the safety of civilian airspace, and will itself investigate all serious LOS incidents.

Safety message

Aircraft separation is a complex operation with many levels of defences to avoid errors and to safely manage the results of the errors that will inevitably be made from time to time by air traffic controllers and pilots. The defences ensure that even if a LOS does occur, the chance of an aircraft collision is still very remote. Safety could be enhanced through understanding and addressing the reasons for the disproportionate rate of LOS occurrences involving civil aircraft in military airspace, and through the ongoing monitoring and investigation of all LOS incidents in civil airspace.

On 17 February 2012, at about 1315 Central Standard Time, a Cessna 210, registered VH-RQD (RQD), was backtracking runway 18 for the general aviation parking area at Darwin airport, Northern Territory.  At the same time, a Cessna 310, registered VH-XXT (XXT), was holding at taxiway Echo 2 (E2) for an intersection departure from runway 29. 

The surface movement controller instructed RQD to cross runway 29 at about the same time as the aerodrome controller cleared XXT for take-off on runway 29. It was estimated both visually and by radar that XXT then overflew RQD crossing the runway by between 150 ft and 500 ft.

Runway incursions are recognised as an ongoing safety concern for the aviation industry and have been cited in numerous accidents world-wide. They can be the result of many different factors and involve pilots, controllers and vehicle drivers. This incident highlights the need for controllers to remain vigilant in monitoring and scanning the runway, both prior to, and after issuing take-off and runway crossing clearances to pilots.

Aviation Short Investigation Bulletin – Issue 17

What happened

On 16 February 2012, a Government Aircraft Factory N22C, registered VH‑ATO (ATO), was conducting a survey flight north-east of Melbourne Airport, Victoria when it was involved in two losses of separation with two Boeing 737 aircraft. The 737s, registered VH‑VZA (VZA) and VH‑TJY (TJY), were on approach to runway 27 at Melbourne Airport. All of the aircraft were in receipt of an air traffic control service.

What the ATSB found

Although air traffic control intended that the pilot of ATO should keep his aircraft separate from VZA and TJY using visual observation (visual (pilot) separation), not all the prerequisites for using this separation method were met. As a result, the onus remained on air traffic control to maintain either a vertical separation standard of 1,000 ft or a surveillance separation standard of 3 NM (5.6 km). Losses of separation occurred when separation between ATO and VZA reduced to 1.5 NM (2.8 km) and 300 ft, and between ATO and TJY when separation reduced to 1.7 NM (3.2 km) and 700 ft.

The ATSB also identified that the controller was not aware of how specific aspects of the survey operation were to be conducted and therefore could not be assured that separation would be maintained.

Additionally, the ATSB identified that limited guidance and training was provided to controllers operating outside the towered environment in the application of the visual (pilot) separation standard, increasing the risk of the incorrect application of the standard.

What's been done as a result

In response to these occurrences, Airservices Australia amended the Manual of Air Traffic Services to include a number of factors to be considered by air traffic controllers prior to assigning visual (pilot) separation. In addition, the training package for Melbourne Airport arrivals controllers was reviewed and updated to ensure that visual separation, transfer of separation responsibilities and associated phraseologies were adequately addressed. Finally, the controller checking regime was strengthened to ensure that any identified training and/or knowledge deficiencies in relation to assigning visual separation, and their impact on controller performance, were addressed.

Safety message

The use of visual (pilot) separation transfers responsibility for separation between aircraft operating in controlled airspace from the air traffic controller to the flight crew of one of the aircraft. This occurrence highlights that, for the standard to be effective, all parties must have a shared understanding. Correct application of the visual (pilot) separation standard ensures that all involved flight crew are aware of their responsibility, thereby assuring that aircraft will not come into unsafe proximity.

At 1310 Western Standard Time on 25 January 2012, a Bombardier Inc DHC-8-202 (Dash 8), registered VH-ZZI (ZZI), was on descent into Broome Airport. At about 17 NM, the flight crew of ZZI advised Broome Tower that they were descending in response to a Traffic Collision Avoidance System (TCAS) Resolution Advisory (RA) alert.

The first officer on ZZI sighted a Pilatus Aircraft Ltd PC-12/45 (PC12) about 1 NM ahead and reported that it passed about 200 to 300 ft to the right and slightly above them.

The traffic was identified as a PC-12, registered VH-MWO (MWO), that had departed Broome Airport at 1300. The pilot of MWO only became aware of the breakdown of separation after ZZI reported a TCAS Traffic Alert (TA), at which time he observed the Dash 8 on his TCAS, behind and on a reciprocal heading. The pilot did not recall hearing a TCAS audible alert.

Both aircraft were operating under Instrument Flight Rules (IFR) and were within airspace controlled by Broome Tower which utilised procedural separation.

The pilot of MWO had not selected automatic direction finding (ADF) as one of the active navigation aids in the aircraft's Electronic Flight Instrumentation System (EFIS). As a result, upon programming new information into the EFIS after departure, MWO unintentionally deviated from the desired outbound track and conflicted with the inbound track of ZZI.

Following the incident, the operator of MWO checked all aircraft in its fleet to ensure that the audio level of the TCAS could be heard above engine noise and radio traffic. MWO was the only aircraft that required the audio level to be increased.

This incident highlights the need for pilots to cross check aircraft navigation performance to ensure accurate track keeping particularly when operating in a procedural separation environment. The incident also highlighted the need to comply promptly with TCAS advisory alerts.

What happened

On 18 January 2012at 2125 Eastern Daylight-saving Time, there was a loss of separation (LOS) between an Airbus A320, registered 9V-TAZ, and an Airbus A340, registered A6-EHH, 907 km NW of Karratha, Western Australia. The aircraft were under the air traffic control of Airservices Australia. The A320 was southbound at FL 350 and the A340 was heading west and cleared to operate in a ‘block’ level, anywhere between FL 340 and FL 360. The aircraft were estimated to cross waypoint TANEM within 2 minutes of each other. The relevant separation standards were 1,000 ft vertical separation or 15 minutes lateral separation at the same position. Controller 1, who approved the block level clearance, did not detect the traffic confliction prior to handing over to controller 2. After a short break, controller 2 handed back to controller 1, and the confliction was detected by controller 2 during the handover. Compromised separation recovery techniques were applied to re-establish vertical separation.

What the ATSB found

The two controllers were experiencing a high workload due to a range of factors, including traffic levels, weather diversions and the airspace configuration, and controller 1 had limited opportunity to consolidate their training and skills before being rostered onto more complex sectors and situations. The ATSB found that Airservices’ processes for monitoring and managing controller workloads did not ensure that newly-endorsed controllers had sufficient skills and techniques to manage the high workload situations to which they were exposed. In addition, Airservices had limited formal guidance regarding how to determine appropriate consolidation periods for en route controllers on one sector before they were transitioned to commence training on another sector. Further safety issues were also identified relating to the application of block level clearances, and the continuing absence of an automated air traffic conflict detection system available for conflictions involving aircraft that were not subject to radar or ADS-B surveillance services.

What's been done as a result

Airservices reported that changes had been made to the configuration, training and rostering arrangements for the airspace sectors involved in the occurrence. The commissioning of a radar in northern West Australia had reportedly alleviated controller workload in two of the airspace sectors, with enhanced surveillance of the majority of aircraft operating in that airspace expected with the mandate of automatic dependant surveillance- broadcast (ADS-B) effective in December 2013. A working group had also been established to determine a suitable workload model to monitor and forecast controller workload on a sector by sector basis. In addition, the first stage of a flight plan conflict function had also been deployed in Brisbane Upper Airspace, with further roll out planned in Melbourne Centre in 2014. The ATSB is not satisfied that Airservices has adequately addressed the identified safety issues regarding formal guidance for consolidation periods for newly-endorsed controllers and the limited formal guidance to controllers and pilots regarding the conditions in which it was safe and appropriate to use block levels. As a result, the ATSB has made formal recommendations to Airservices.

Safety message

Ideally the best way of managing workload is to reduce the level of work demands and distractions. If the work demands cannot be reduced, then another option is to ensure the controllers have the experience, skills and techniques to effectively manage their task demands. Overall, high workload can have significant effects on a controller’s performance, and it needs to be monitored and managed using a systemic approach, particularly for less experienced controllers but also controllers who have recently received a new endorsement. The instances of other recent loss of separation occurrences involving high workloads and newly-endorsed controllers on other sector indicates that this problem was not restricted to the sectors involved in this occurrence.