DH8B, Kangerlussuaq Greenland, 2017
From SKYbrary Wiki
|On 2 March 2017, a DHC8-200 took off from Kangerlussuaq in normal day visibility without clearance and almost immediately overflew three snow clearance vehicles on the runway. The Investigation identified a number of likely contributory factors including a one hour departure delay which the crew were keen to reduce in order to remain within their maximum allowable duty period and their inability to initially see the vehicles because of the runway down slope. No evidence of crew fatigue was found; it was noted that the vehicles involved had been in contact with TWR on a separate frequency using the local language.|
|Actual or Potential
|Air-Ground Communication, Human Factors, Runway Incursion|
|Aircraft||BOMBARDIER Dash 8 Q200|
|Type of Flight||Public Transport (Passenger)|
|Intended Destination||Sisimiut Airport|
|Take off Commenced||Yes|
|Flight Phase||Take Off|
|Location - Airport|
Inadequate Aircraft Operator Procedures,
|Tag(s)||Take off without clearance|
|Tag(s)||Procedural non compliance,|
|Tag(s)||Accepted ATC Clearance not followed,|
Incursion pre Take off,
No Single Runway Occupancy Frequency
|Damage or injury||No|
|Causal Factor Group(s)|
On 2 March 2017, a Bombardier DHC8-200 (OY-GRO) being operated by Air Greenland on a scheduled domestic passenger flight from Kangerlussuaq to Sisimut departed from an active runway in day VMC without having received a take-off clearance and was just able to clear three snow clearing vehicles which the crew sighted ahead during their take-off roll.
An Investigation into this Serious Incident was carried out by the Danish Accident Investigation Board. Data from the QAR was retrieved but relevant data from the SSCVR had been overwritten by the time the event was reported to the Board by Kangerlussuaq ATC a little under 2 hours after it had occurred.
The 47 year-old Captain had 14,800 hours total flying experience; the experience on type was not stated. The 44 year-old First Officer had 3,494 hours total flying experience which included 2,681 hours on the aircraft type involved. The investigated event occurred during a seven sector rostered flight duty; the position of the investigated flight in that sequence was not stated but an investigation as to whether fatigue may have been a factor in the performance of either pilot found no evidence of this.
It was established that after clearance to line up on runway 27 ready for departure, the departing aircraft had read this clearance back and then taxied from the apron via taxiway 'A' onto the runway. A take-off clearance was not given as the TWR, where a trainee controller was in position and being overseen by an OJTI, had, in anticipation of the departure of the DHC8, instructed three vehicles involved in snow clearance on the runway to vacate via taxiway 'D' using "the local FM channel" and communicating in the local language. This local frequency was not promulgated in the AIP.
Just seven seconds after arriving as cleared in the 27 take-off position at taxiway 'A' and after the crew had "been in doubt whether or not a take-off clearance had been issued but (had) agreed that the aircraft was cleared for take-off", the aircraft began its take-off roll. An ASDA/TODA of 2579 metres was available from that point and the aircraft was operating at significantly below its permitted MTOM. The crew had no visual contact with the vehicles because they were located on the western end of the runway where the last 923 metres, beginning just beyond taxiway 'D', slopes down to a threshold elevation which is 65 feet lower than the opposite end of the runway (see the runway longitudinal profile below). It is not made clear whether the crew subsequently did sight any of the vehicles later in the take-off roll.
As the aircraft passed taxiway 'C', (see the annotated aerial view above), the first vehicle was about to vacate the runway as cleared at taxiway 'D' but the other two still had around 500 metres to go before reaching that point. The take-off was continued and the aircraft became airborne approximately 35 seconds after starting to roll, subsequently passing approximately 300 feet overhead of the nearest vehicle. It was estimated that the closest the aircraft had come to any of the vehicles before becoming airborne was approximately 900 metres. It was reported that "the student air traffic controller at Sondrestrom TWR observed that the aircraft had started its take-off roll and pointed it out to the instructor air traffic controller" and that "for a short while, the instructor air traffic controller considered an instruction on (instructing a rejected) take-off but had (by then) noticed that the aircraft had started rotating".
It was noted that as it taxied out, the aircraft was running about an hour behind schedule and the Investigation considered that a relevant context for the imminent departure could have been that the flight crew were "eagerly trying to catch up on time in order not to exceed their maximum daily duty period of 10:30 hours (seven sectors)" and that, this "might have mentally provoked them to act from their expectations rather than the factual reality".
- Motivational judgement
- Goal conflicts
- Workload and stress.
The Investigation also considered the absence of what were described as "soft safety barriers" including:
- The aircraft operator’s SOP for the use of aircraft external lights on the ground did not meet the recommendations to that effect aimed at maximising aircraft conspicuity contained in the ICAO 'Manual on the Prevention of Runway Incursions' Doc 9870.
- The ICAO PANS-ATM Doc 4444 advice to consider adding "(and) wait" to line up clearances, which were not also issued as take-off clearances in appropriate circumstances, had not resulted in this action being taken in this case. The Investigation considered that the addition of "and wait" would have made the line up clearance more "attention demanding".
- TWR did not re-transmit ATC voice communications between the TWR and the runway snow clearing vehicles to aircraft in the air or on the ground. It was considered that whilst "there is delicate balance between securing situational awareness on the one hand and balanced and not least relevant ATC voice communication on the other hand" there were safety-related weaknesses in some of the phraseology used and in dual language and dual frequency use.
The Conclusion of the Investigation was that the aircraft "departed from an engaged runway at Kangerlussuaq without an ATC clearance".
Four Contributory Factors were identified as having "influenced the sequence of events" as follows:
- The actual flight was approximately one hour behind the scheduled time of departure.
- An over-motivated flight crew acted from their expectations rather than the factual reality.
- Runway down slope prevented the flight crew from spotting the three runway snow clearing vehicles.
- Soft safety barriers like the applied ATC phraseology and the flight crew Standard Operating Procedures on use of landing lights did not provoke or raise the attention of the flight crew in such a way that their decision making process changed.
Relevant Safety Action taken by the aircraft operator in response to the event during the Investigation was noted to have included the following:
- Flight crew were reminded of their responsibility to secure the SSCVR for safety investigation purposes.
- A new SOP was "defined" by which both pilots must confirm ATC clearances with each other.
- A new SOP for the use of landing, taxi and wing inspection lights was introduced.
The Final Report was published on 11 May 2017. The Board did not issue any Safety Recommendations "because the Investigation did not reveal systemic flight safety lapses".
- Loss of Separation
- Runway Incursion
- The Human Factors "Dirty Dozen"
- Confirmation Bias
- Commercial Pressures
- Expectation of Clearance
- External Lights
- Crew Resource Management
- Cockpit Voice Recorder (CVR)
- ATC Radio Use by Airside Vehicles
- Multi-language ATC Operations