E120, Amsterdam Netherlands, 2016
From SKYbrary Wiki
|On 18 January 2016, an Embraer 120 crew made a night takeoff from Amsterdam Runway 24 unaware that the aircraft was aligned with the right side runway edge lights. After completion of an uneventful flight, holes in the right side fuselage and damage to the right side propeller blades, the latter including wire embedded in a blade leading edge, were found. The Investigation concluded that poor visual cues guiding aircraft onto the runway at the intersection concerned were conducive to pilot error and noted that despite ATS awareness of intersection takeoff risks, no corresponding risk mitigation had been undertaken.|
|Actual or Potential
|Human Factors, Runway Excursion|
|Type of Flight||Public Transport (Cargo)|
|Origin||Amsterdam Airport Schiphol|
|Take off Commenced||Yes|
|Flight Phase||Take Off|
|Location - Airport|
|Airport||Amsterdam Airport Schiphol|
Ineffective Regulatory Oversight,
Inadequate Airport Procedures
|Tag(s)||Misaligned take off|
|Damage or injury||Yes|
|Causal Factor Group(s)|
Air Traffic Management,
|Group(s)||Air Traffic Management,|
On 18 January 2016, an Embraer 120 (EC-JBD) being operated by Swiftair on an international scheduled cargo flight from Amsterdam to London Stansted was found, after completion of its flight, to have damage to both the right propeller and the right side of the fuselage with this damage considered indicative of a runway edge takeoff that had been made in night VMC. Damage to both nose wheel tyres and to a number of runway edge lights was subsequently found. The crew stated that they had been unaware of their error.
An Investigation was carried out by the Dutch Safety Board (DSB). Relevant data were downloaded from the aircraft FDR and CVR and available from the airport surface movement radar at a resolution greater than available to the runway controller in real time.
It was noted that the Captain had a total of 3,653 flying hours including 1,802 hours on type and was a Line Training Captain for Swiftair. The First Officer had a total of 1,510 hours which included 1,300 hours on type. No evidence was found that pilot fatigue had played any part in the event.
It was established that the taxi clearance initially given to the flight for a runway 24 departure had been to join the parallel taxiway A and proceed to the holding point on S7. After taxiing had commenced, the controller had asked the crew if they would like an intersection takeoff and when they replied in the affirmative, the controller had offered either S4 or S5 and after a preference for S5 had been expressed, the corresponding re-clearance was given. Although the First Officer had been designated as PF for the sector, the absence of a steering tiller at their seating position meant that until the aircraft was lined up on the runway, the Captain would be taxiing the aircraft and the First Officer acting temporarily as PM.
As the aircraft approached the turn onto taxiway S5, the First Officer acknowledged a change from GND to TWR and checked in with TWR. As the aircraft made the turn onto S5, the TWR controller gave takeoff clearance and as the aircraft was turned right to align with the right side runway edge lighting without crew awareness, TWR instructed the aircraft to make a right turn onto a heading of 270° after departure. Once lined up, control was passed to the First Officer and it was the Captain who acknowledged the heading instruction. It was found that 20 seconds elapsed between the receipt of the heading instruction and the setting of takeoff power during which time there was no conversation between the pilots except for that consequent upon completing the Takeoff Checklist.
After engine shutdown on completion of the planned flight the crew were alerted by hand signal from ground personnel that there was visible damage to the aircraft. Their subsequent inspection identified damage to the right side of the fuselage and to three blades of the right propeller, one of which had a metal wire embedded in it. The following morning, a formal inspection of the aircraft was made by the UK AAIB on request from the DSB and confirmed that the damage was confined to the right side of the aircraft. Several puncture marks and one puncture hole were found in the fuselage all associated with yellow paint marks contrasting with the white-painted fuselage. Damage to three right propeller blades was also found with yellow paint associated with leading edge damage to one associated with “yellow paint scuff marks”. Both nose wheel tyres were also found to be damaged with yellow paint marks visible with one also partially deflated and detached from its rim.
Shortly after the aircraft had completed its flight, a bird controller carrying out a routine inspection of runway 06/24 found three right side edge lights near to taxiway S5 had been destroyed and four others further down the runway were broken. A more detailed inspection of these damaged/destroyed yellow-painted lights found that at least one of them had black marks likely to be attributable to aircraft tyre contact.
CVR data recorded the edge lighting impacts and also that after the first of these, the crew had discussed “a sound” but attributed it to a falling pilot briefcase. Both pilots subsequently acknowledged during interview that “some bumps” had been felt during the takeoff roll but stated that they had been unsure whether the source was the nose wheel or the main wheels and noted that such bumps were not unusual as the nose gear ran directly over the runway centre line lights.
CVR data also showed that an aural warning of a stall warning system fault had been activated when the aircraft was pitched up for takeoff and had been correctly assessed as false at the time of occurrence. The Investigation noted that this warning would occur in the event of any detected discrepancy between the two angle-of-attack vanes and that there had been no annunciation of a stall warning or any evidence from the FDR that conditions which would have given rise to one had existed. In the absence of any repetition or continued system malfunction, it was therefore concluded that the most likely explanation of the fault indication was that some of the pieces of the impact-damaged runway edge lights which had been directed into the propeller arc had then been catapulted in various directions causing not only fuselage impacts but potentially an impact with the right hand angle-of-attack vane which might have momentarily compromised its correct position.
It was noted that 3500 metre-long runway 24/06 is 45 metres wide and has an additional 15 metre-wide shoulder on both sides. The exact layout of the S5 runway access, primarily designed to facilitate runway crossing to/from S8 on the opposite side and being one of 7 access points on the northwest side of the runway, was examined - see the illustrations below – and was nominally an “exit taxiway”. It was noted that alignment with runway 24 from S5 required a 120° right turn. It was also found that this turn was not provided with a corresponding extension of the taxiway centreline in contrast to the 30° turn onto the runway from S6 and the 90° turn onto the runway from S7E. All marked access runway routes in the vicinity of S5 were also provided with centreline lighting.
The absence of taxiway lighting was noted to be contrary to the normally applicable EASA requirements which require that this should be provided on any taxiway intended for use at night in runway visual range conditions of 350 metres or more “and particularly on complex taxiway intersections and exit taxiways”. The same guidance was also noted to state that such lighting is not necessary “on exit taxiways where the traffic density is light and taxiway edge lights and centre line marking provide adequate guidance” but this exception was not valid for S5 since centreline marking to/from the runway was not provided. The Investigation also considered that the airport operator’s view that S5 was covered by the EASA “light traffic density” exception noted above was inappropriate and referenced the ICAO Annex 14 definition of aerodrome traffic density (not used by EASA) in support of its opinion that “the airport as a whole should be considered as a heavy traffic density zone”.
The ANSPs procedures for intersection departures were examined and it was found that despite the absence of both a centreline marking continuing onto the runway and any centreline lighting, S5 was included in the list of “recommended intersections” available for use by aircraft departing from runway 24. It was also noted that the ANSP (LVNL) “believed that it was not relevant” for air traffic controllers to know exactly which runway entries were available or what lighting was available at them given that their Operations Manual specifies which entries may be used in what conditions.
In the context of the “known risks associated with intersection takeoffs” the Investigation noted that the LVNL ‘Procedural Handbook’ states that the runway controller must only allow jet aircraft to take off via an intersection if there is an operational reason to do so. Examples of such operational reasons are given as changing the departure sequence to increase takeoff capacity, avoiding jet blast affecting traffic on another runway and avoiding crossings of active runways. In the case of the investigated event, LVNL stated that the first of the above-quoted operational reasons had applied. However, it was noted that LVNL's Operations Manual content on reasons for offering an intersection takeoff was not applicable to propeller-driven aircraft and also observed that it “did not include any stipulations focusing on reducing known risks that exist when offering or permitting intersection takeoffs”. It was also noted that although the LNVL's internal investigation into the event stated that it had been “focused on what LVNL could possibly do to prevent any repetition of this kind of incident”, it limited itself to “bringing the results of the investigation to the attention of the local Runway Safety Team” (RST). It was noted that the RST meeting at which the report of LVNL's internal investigation was subsequently considered decided only to add a runway lead-on extension to the existing S5 taxiway centreline and provision of centreline lighting for S5 was not progressed.
The report of the internal investigation into the event carried out by the Airport Operator was also examined and it was concluded that its objective “namely to determine the extent to which the infrastructure, the current procedures at Schiphol Airport and the actions of the airport party or parties involved played a role in the occurrence”, had not been achieved. The report was found to state that “no clear conclusions could be drawn and therefore no recommendations could be made”.
The Investigation also noted that in response to a DSB Safety Recommendation following its investigation of a 2010 taxiway takeoff at Amsterdam, a risk assessment conducted by the collaborative ‘Schiphol Safety Platform’ (VpS) had identified requesting/offering an intersection takeoff as one of the top five risks during the pre-takeoff phase of departures, “there were no further conclusions, actions or proposals for mitigating measures associated with this risk assessment”.
Previous Similar Events
It was noted that there had previously been similar occurrences to that under investigation at both Amsterdam and, in similar circumstances, more widely. In October 2014, an almost identical event in which an aircraft had also lined-up on the right hand edge of runway 24 after entering via intersection S5 had occurred but because the error had been seen and reported by the pilot of another nearby aircraft, the situation was corrected. No risk mitigation action was taken by LVNL after this event yet it was noted during the Investigation that four of six “runway misalignment risk factors” identified in the June 2010 report of the ATSB study Factors Influencing Misaligned Take-off Occurrences at Night had been present. Investigations into similar events elsewhere since the publication of the ATSB report were noted as having included a 2013 event at Auckland New Zealand and a 2014 event at Biggin Hill UK.
The occurrence of these and many other previous events was considered to have been made more likely because pilot situational awareness can be compromised by the difficulty of distinguishing edge lights from centreline lights when they are viewed along the runway axis given that both are white and the distinction depends on “the pattern of edge lights and the relationship of this pattern with other lights and visual cues”. The Investigation noted that “if individual edge lights could be identified as such directly, rather than through a process of interpretation”, it would be easier for pilots to detect an incorrect line up; it further observed that “modern lighting technology offers more options to identify lights directly than the tungsten lighting technology on which the current standards are based”.
The Conclusions of the Investigation included the following:
- During takeoff, the crew interpreted the right-hand side runway edge lights as the runway centreline lights.
- Several runway edge lights were struck by the nose landing gear and it is suspected that debris from a number of these lights caused the aircraft damage subsequently found.
- The large turning angle required to align the aircraft with the runway centreline when entering at intersection S5 in combination with the discontinuity of the taxiway S5 centreline and absence of any taxiway centre line lighting contributed to the misalignment.
- The absence of a continuous centre line marking on taxiway S5 leading to the centre line of Runway 06/24 was not compliant with the corresponding EASA specifications.
- The absence of centreline lighting on taxiway S5 was not compliant with the corresponding EASA specifications.
- the issue of an ATC takeoff clearance during the turn onto S5 and the subsequent issue of a departure heading as the aircraft turned onto Runway 24 might have distracted the crew.
- The ANSP LVNL was already aware of the operational risks associated with intersection takeoffs before the investigated event had occurred but had not taken any operational measures to mitigate those risks.
- The ANSP LVNL continued to identify intersection S5 as a recommended intersection at night despite the fact that the intersection did not have centre line lighting and did not consider the question of whether intersection S5 was justifiably designated as a ‘recommended intersection’ in the Operations Manual and if offering the intersection concerned was a wise choice. It also did not reconsider its position after the decision was taken not to install centre line lights on intersection S5.
Three Safety Recommendations were made as a result of the Investigation as follows:
- that the International Civil Aviation Organisation (ICAO) initiates the process to develop a Standard for runway edge lights in Annex 14 Volume 1 ‘Aerodrome Design and Operations’ that would allow pilots to identify them directly without reference to other lights or other airfield features.
- that the Amsterdam Airport Operator proactively takes measures, in line with the above recommendation to ICAO, which prevent pilots from interpreting the runway edge lights as the runway centre line lights.
- that LVNL (the Netherlands ANSP) only allows air traffic access to the runway for takeoff using intersections outside of the uniform daylight period when such intersections are equipped with centreline lighting.
The Final Report of the Investigation was published on 2 November 2018.
- Runway Excursion
- Risk Management
- Risk Mitigation
- Taxiway Lighting
- Runway Lighting
- Taxiway Surface Markings and Signs
- B733, Amsterdam Netherlands, 2010
- A343, Auckland New Zealand, 2013
- GLF3, Biggin Hill UK, 2014
- Factors influencing misaligned take-off occurrences at night, ATSB Australia, 2010