Description

On 27 July 2021, an Airbus A320 (9V-JSM) being operated by JetStar Asia on a scheduled international passenger flight from Singapore Changi to Jakarta Soekarno-Hatta took off uneventfully from runway 20L in normal night visibility despite crew recognition that some runway lighting was not on. ATC were not advised of this by the crew until after the flight was airborne after which all lighting was selected on.

Investigation

Once advised of the event more than 24 hours after it had occurred, the Singapore Transport Safety Investigation Bureau (TSIB) began an Investigation into the circumstances. By this time, relevant data from the aircraft CVR had been overwritten but relevant CCTV footage was available. 

The Flight Crew

The 42 year-old Captain who was acting as PF for the flight had a total of 12,193 hours flying experience all but 229 hours of which were on type and he had flown 105 hours in the preceding 90 days. The 50 year-old First Officer with him had a total of 6,706 hours flying experience of which 4,298 hours were on type. The 34 year-old controller involved had been rated to operate at Changi for almost six years.

What Happened

As the aircraft was taking the taxi route shown in the illustration below and shortly after the change from GND to TWR frequency, the flight was issued with a takeoff clearance from runway 20L (‘runway 3’) by the Changi East Tower controller as it approached the holding point. Once lined up, the crew stated that the edge lights had been on and since centreline lighting was not a company requirement for night takeoff, a rolling takeoff in visibility recorded as greater than 10 km had followed. Company night takeoff requirements also included that runway end lights were illuminated but this was not checked. 

The taxi out route taken by the aircraft (not to scale). [Reproduced from the Official Report]

It was noted that the First Officer had subsequently stated that he had considered suggesting to the Captain that the absence of centreline lighting should be queried with ATC before taking off but had instead “waited to see what the Captain would do” after which the opportunity to do so had passed. 

After becoming airborne, the crew noticed that the edge lighting was actually off and advised this and the absence of centreline lighting to the controller who then switched on the runway lighting, all of which had, confirmed by CCTV, previously been off.

Why It Happened 

The operator’s procedures required that the runway end lights must be illuminated for a night takeoff but this was not verified. The controller had forgotten to switch on the runway lighting and this omission was not then detected. It was found that controllers’ judgment of what constituted “the hours of darkness” as a basis for deciding when to switch on aerodrome lighting for night operations could be very subjective which could result in lighting systems within different areas of responsibility being switched on at different times. It was also assessed likely that expectation bias had provided the basis on which the controller had concluded that the runway lights were all on and also the basis on which the pilots concluded that the runway edge lights were on.

Discussion 

• The pilots involved were both based in Singapore and familiar with operations from the airport and could therefore have been reasonably expected to regard runway use at night with the centreline lighting off to have been unusual and worthy of a query to ATC prior to takeoff if encountered. It was also noted that despite the fact that operator procedures required the runway end lights as well the runway edge lights to be on for night takeoffs, this was not been discussed. 
• Controllers were expected to use their judgement as to when to switch on taxiway and runway lighting during the day/night transition so that they were on during “the hours of darkness”. At the time of the takeoff, the moon had not risen and sunset had occurred 55 minutes prior to the takeoff. The corresponding taxiway lighting had been selected on 34 minutes prior to sunset but the controller had then forgotten to switch on the runway lighting. 
• The installed lighting on runway 02L/20R was as follows:
  • Bi-directional edge lighting with the beams aligned with the runway edges (this was different to the edge lighting installed on runways 02L/20R and on 02C/20C which was omni-directional and could therefore be seen from the control tower. 
  • Bi-directional centreline lighting 
  • Runway end lighting 
• The effect of bi-directional edge and centreline lighting was such that controllers in the Changi East Tower (CET) would be unable to see this lighting within approximately 500 metres either side of their TWR cabin (see the illustration below). It was, however noted that controllers “may see those lights that are further away from this section” although they “may not be easily discernible”.

The field of view of the runway edge lights from the CET (not to scale). [Reproduced from the Official Report]

• A taxi trial with the same aircraft type stationary at the threshold of runway 20L and the centreline and runway edge lights off showed that it was unlikely that:
  • the runway 20L edge lights could appear ON when they were OFF; 
  • the blue taxiway edge lights near the threshold of runway 20L could be misidentified as white runway edge lights;
  • the lights from the airport vicinity could have been mistaken as runway edge lights. 
• The routine assessment of the runway 02R/20L lighting system prior to the opening of this new runway in November 2020 was found to have included flight checks to confirm the suitability of the runway lights for flight operations but there was no requirement to consider whether the edge lights could be seen - or could be easily seen - by controllers in the CET. It was found that a cursory view from the CET of the centre portion of the edge and centreline lights at night was similar whether the lights were on or off and this may not have been recognised by controllers during the relatively short period since the runway had opened.
• In conjunction with the opening of the new runway, a new A-SMGCS had replaced the previously used Airfield Ground Lighting Control and Monitoring System as the means to provide control and indication of airfield lights to controllers by integrating lighting control/indication and the positions of aircraft and vehicles into a single system and display. However, it was found during the Investigation that just a quick glance at the display by a controller would not allow them to discern whether the runway lights were on or off and that they would need to look more closely at the display.
• Although the runway lighting system was in full compliance with ICAO standards, the risk assessment apparently failed to consider the difference between the omni-directional lights installed on the two original runways and the bi-directional lights on the new runway. It therefore did not consider whether the on/off status of the new runway’s edge lights could be easily determined by visual reference or establish a method to achieve this.   
• There had been no controller feedback to the ANSP about any difficulty in discerning runway 20L lighting status on the A-SMGCS display despite the existence of multiple reporting channels including a confidential voluntary internal hazard reporting system and an industry-wide aviation confidential reporting system. It was found that controllers did not appear to have recognised “the safety hazard” which their inability to see the bi-directional edge and centreline lights at the central part of runway 20L represented.

The Conclusions of the Investigations were formal statements based on the above findings.

Safety Action taken during and known to the Investigation as a result of the event was noted to have included the following:

• JetStar Asia as interim action prior to an intended OM amendment, issued a Flight Standing Order to remind pilots of the need to:
  • ensure that runway lighting configuration is correct and complies with the operator’s minimum requirements for that particular runway for night operations and/or low visibility conditions.
  • seek clarification from ATC before takeoff is commenced if the runway lighting configuration is not what is expected.
• The Changi ANSP 
  • required the lead controller at CET to inform the Changi Tower Watch Manager when the aerodrome lights within CET area of responsibility have been switched on/off at sunset/sunrise. 
  • a table of sunrise/sunset timings, extracted from the Singapore AIP, has been provided to controllers. 
  • installed an alert system both at CET and at the Changi TWR which will activate near sunset as a reminder to controllers to switch on the aerodrome lighting.

Three Safety Recommendations were made as follows: 

• that JetStar Asia remind its pilots of the dangers of cognitive biases and of the ways to mitigate the effects of such cognitive biases when checking the status of runway lights. [RA2022-001]
• that the Changi ANSP remind controllers of the dangers of cognitive biases and of the ways to mitigate their effects when checking the status of runway lights. [RA-2022-002]
• that the Changi ANSP enhance the promotional activities of its hazard reporting system to increase its controllers’ awareness of the system. [RA-2022-003]

The Final Report was published on 17 May 2022.

Related Articles

• Runway Lighting
• Bias
• ATC Expectation Bias
• Pilot Expectation Bias
• Risk Assessment