Introduction
On 5 August 2019, during early evening operations at Sydney Airport, three aircraft operated by Qantas Group were using runway 34R in various phases of flight. A de Havilland Canada Dash 8 (QantasLink) was nearing the end of its landing approach, a Boeing 737 (Qantas) was on the final segment of an independent visual approach, and an Airbus A330 (Qantas) was awaiting instructions and clearance to line up and take off, following the MARUB SIX standard instrument departure (SID).
Sequence of Events
The air traffic was managed by an aerodrome controller (ADC) position occupied by a controller in the late stages of training, supervised by an on-the-job-training instructor (OJTI). After the Dash 8 landed and taxied off the runway, the trainee ADC cleared the A330 for an immediate take-off, and the crew complied. Assessing that runway separation between the departing A330 and the arriving 737 might be insufficient, the trainee ADC instructed the 737 crew to go around (execute a missed approach).
The 737 flight crew initiated the missed approach by climbing on runway heading but climbed through the mandatory turn altitude. The aircraft turned when instructed by the trainee ADC. Meanwhile, the A330 followed the SID track, turning right shortly after passing the departure end of the runway. The two flight paths began to converge.
As both aircraft were turning right and climbing, the A330 flight crew received a traffic alert from the onboard traffic collision advisory system. Shortly after, the A330 first officer sighted the 737. The controllers maintained visual contact with both aircraft throughout the sequence. In the absence of compliant methods to separate aircraft at night, the trainee ADC attempted to establish horizontal separation by instructing the 737 crew to turn onto a heading divergent from the A330's outbound track. The separation between the aircraft reduced to about 0.42 NM (800 m) laterally and 508 ft vertically, constituting a loss of separation.
Findings
The Australian Transport Safety Bureau (ATSB) found that the loss of separation and close proximity resulted from a series of events, each minor but collectively significant. The spacing between the 737 and the landing Dash 8 was less than permitted without coordination between controllers. Additionally, the 737's speed during part of its final approach exceeded the approach design specification, and the flight crew did not advise the ADC. These factors increased the risk of compromised runway separation and the associated go-around.
The trainee ADC's mental model of the traffic situation did not fully account for the effects of the 737's delayed and relatively wide turn, and the expected flight path of the A330 was further from the 737 than actually occurred. Consequently, the trainee ADC's actions, while reducing the extent of close proximity, were not optimal. No safety alert or avoiding action advice was given to either flight crew to notify them of their proximity. The trainee ADC also did not modify the A330's projected flight path to increase distance and re-establish separation standards sooner.
The OJTI was not confident that runway separation could be achieved or that turn instructions issued to the 737 were sufficient mitigation, yet did not provide effective prompts or intervene, mindful that the trainee was meant to demonstrate ability without instructor intervention.
The ATSB identified safety issues related to the management of the MARUB SIX SID and the missed approach procedure for runway 34R, which directed aircraft onto similar outbound tracks, potentially requiring controller intervention to maintain separation. In daytime, controllers could vector aircraft, but at night, no procedural controls existed for separation below the minimum vector altitude. Furthermore, compromised separation training scenarios involving aircraft below this altitude at night were absent. Although these issues were known among Sydney controllers and Airservices had addressed similar problems at other airports, they remained unaddressed at Sydney.
Other findings not considered contributory included: the tower shift manager was engaged in a controller function and unaware of the go-around and developing loss of separation until after the event; air traffic control transfer after the occurrence; and the location of a relevant navigation waypoint in the 737's flight management computer was incorrect.
Safety Actions
In 2020, Airservices conducted a risk assessment on the runway 34R missed approach procedure and MARUB SIX SID, and on distances between successive arrivals. Subsequently, Airservices redesigned the missed approach procedures for Sydney's runway 34R to increase the likelihood that distance would be maintained with an aircraft departing on a SID from the same runway. The ATSB urged Airservices to monitor safety outcomes on an ongoing basis.
Also in 2020, Airservices advised that compromised separation scenarios where an aircraft operates below the minimum vector altitude at night would be included in the Sydney tower controller instructor guide. In 2023, Airservices advised that the training program included a missed approach with a preceding departure in instrument meteorological conditions. Airservices also improved risk management processes, issued a standardization directive on spacing for arrivals, established a focus group to understand shared risk factors, and implemented actions to improve the operational availability of tower shift managers. All safety issues identified by the ATSB have been adequately addressed.
Qantas conducted several safety actions, including updating the missed approach coding in its 737 flight management computers, incorporating related scenarios into cyclic training, and updating its flight data analysis program to monitor approach speeds and traffic collision avoidance system data.