On 28 March 2018, a Boeing 787-9 (G-CKWC) being operated by Norwegian on a scheduled international passenger flight from London Gatwick to Buenos Aires at night in VMC was observed by ATC to only just get airborne before reaching the end of the runway. There was no communication between the crew and ATC about this after takeoff and it took some time for the event to be recognised as initially a reportable event and eventually as a Serious Incident.
An Investigation by the UK AAIB was only commenced after it first learned of the event on 14 May 2018 from the UK CAA who themselves had not received a MOR from the aircraft operator detailing the occurrence until 30 April 2018. They in turn had learned of the event only after an internal report raised by ATC had been copied to them and had then evaluated the downloaded flight data against the performance requirements before concluding that it was within MOR reporting requirements. Neither the ANSP nor the airport operator reported the event to the CAA as a MOR and quite apart from MOR reporting requirements, the Investigation noted that, since the event was “a gross failure of the aircraft to achieve its predicted takeoff performance”, there was a statutory requirement to report it as a Serious Incident without delay.
It was noted that the 63 year-old Captain had accumulated a total of 18,765 flying hours which included 699 hours on type. No information was provided about the First Officer's experience or which pilot was designated as PF for the takeoff and the pilots were not interviewed. It was found that although the crew were familiar with the airport and its main runway 26L, their use of the standby runway 26R had been "infrequent".
It was established that the flight was scheduled to depart 15 minutes prior to a planned switch from the main westerly runway to the standby runway with the corresponding NOTAM advising that the last main runway departure would be 5 minutes before the changeover. The crew had therefore begun re-planning for a departure from the standby runway which, because of its shorter length - a TORA of 2,565 metres rather than 3,255 metres - required a reduction in the originally planned cargo load. The revised loading allowed departure from the standby runway using rated thrust.
The re-planning proved well judged as the aircraft did not leave its gate until it was too late to use the main runway. The aircraft was cleared to taxi to holding point P1 on taxiway P and this was followed by clearance to line up on runway 26R after a landing Airbus A320. Once past the holding point, this involved crossing taxiway J and then making a right turn onto taxiway AN which then led directly to the beginning of runway 26R without any change of direction (see the illustration below). The beginning of the runway was marked by a transverse white line and an illuminated sign on the left side giving the TODA from that point.
The area at the beginning of runway 26R showing the position of the TODA sign. [Reproduced from the Official Report]
Whilst on taxiway AN, the crew received their takeoff clearance but then failed to recognise where the beginning of the runway was or notice the TODA sign and continued taxi speed for 417 metres until they reached the displaced landing threshold. As the aircraft reached this point, a rolling takeoff was initiated with only a 2140 metre TORA now remaining. The TWR controller observed that the aircraft did not appear to rotate until approximately 600 metres before the end of the runway. The crew subsequently reported that although the takeoff "seemed normal", it had used the full length of the runway and both pilots had commented after getting airborne that "there was not much runway remaining at lift-off".
It was noted that the initial part of runway 26R prior to the displaced landing threshold had red rather than white edge lighting and no centreline lighting which corresponded to the applicable regulatory requirements. It also included one bar and some centre lights of the approach lighting system. The displaced threshold had green threshold lights as well as a set of Runway Threshold Identification Lights (RTILS) but the latter were angled such that they were not intended to be visible to an aircraft on the ground. The contrast between the relatively dark area of the runway prior to the displaced landing threshold and that of the rest of the runway was evident from comparative photographs representative of the view from the flight deck taken during the Investigation.
It was also noted that ever since the introduction of the northern standby runway in the mid 1980s, there had been a history of crews misidentifying the start of runway 26R which was originally only a parallel taxiway serving the main runway and was still operated as such most of the time. An early response to the misidentification problem was the installation in the early 1990s of the TODA sign at the beginning of the runway. This sign is visible only when the standby runway is in use as an alternative to the main runway - the two runways cannot be in use as such at the same time due to close spacing. However, similar incidents have continued to occur and the Investigation noted that in the 6 months prior to the takeoff being investigated, ATC had recorded at least four other incidents in which aircraft of other operators had not started their takeoff roll at the beginning of the runway 26R TORA.
Takeoff Performance Issues
A review confirmed that in the prevailing weather conditions, the 223,813 kg takeoff weight of the aircraft required an ASDA of 2,564 metres which was the same as the ASDA for the full length of Runway 26R, whereas from the point where the takeoff was actually commenced, the displaced landing threshold, the ASDA was only 2,156 metres. Had an engine failed during the takeoff as conducted prior to reaching V1, a rejected takeoff would have been likely and the overrun may have continued beyond the airport boundary which was only around 2,250 metres from the beginning of the takeoff. It was calculated that the aircraft had been about 12 tonnes too heavy for the takeoff made to satisfy regulated takeoff performance criteria. It was also noted that “the TODA was also insufficient to meet regulatory requirements for obstacle clearance should the aircraft have continued the takeoff after an engine failure at V1”. However, it was noted that some of the assumptions which are necessarily made when calculating regulated takeoff performance are necessarily conservative. In this specific case, the use of wet runway data when the runway was only damp, the fact that the runway was grooved and the rolling takeoff would all have had a beneficial effect relative to calculated takeoff performance had an engine failure occurred near to V1.
The formally stated Conclusion of the Investigation was as follows:
The aircraft began its takeoff roll from the displaced threshold of Runway 26R rather than the beginning of the runway. The crew did not identify the beginning of the runway and instead taxied the aircraft forward to the landing threshold. A combination of an unusual straight-line runway entry, a perceived lack of lighting in the pre-threshold area and the bright threshold lights ahead contributed to the crew not identifying the beginning of the runway.
From the point at which the aircraft began its takeoff roll, its performance did not meet regulatory requirements for both stopping and continuing should an engine have failed close to V1. The risks in both cases were significant to the aircraft and its occupants.
Safety Action taken by the Airport Operator and/or the ANSP following this Serious Incident and known to the Investigation included but was not limited to:
- The ANSP committed to raising awareness with their staff about their obligations to report Serious Incidents to the AAIB under a process which had, during this Investigation, already been formally clarified as a result of a similar failure to report by UK ANSP personnel in respect of another takeoff performance-related Serious Incident at Belfast International.
- The amendment of the standard NOTAM used to promulgate the closure of the main runway and the use of standby runway 08L/26R to include a more explicit description of the position at which the takeoff roll should commence.
- After a review of the arrow markings on the standby runway before the displaced landing threshold on both 08L and 26R which found that they were not in compliance with applicable regulations in that they were too widely spaced, this deficiency was remedied by increasing the number of such arrows from 5 to 8.
The Final Report of the Investigation was published on 6 December 2018. No Safety Recommendations were made.