On 2 January 2015, a Saab 340B (G-LGNL) being operated on a domestic scheduled passenger flight from Stornoway to Glasgow left the runway at speed during a daylight take off in normal visibility and continued for 250 metres eventually coming to a stop on the grass some 35 metres from the edge of the runway. After a delay in the subsequent passenger-initiated emergency evacuation because the propellers were still turning, passengers evacuated via the right hand over wing exit. One of the 29 passengers was found to have sustained a serious injury. Extensive damage was sustained to the nose landing gear, both power plants and the underside of the aircraft.
The aircraft track during the excursion and its final stopping position (reproduced from the Official Report)
An Investigation was carried out by the UK AAIB assisted by data recovered from the 2 hour CVR and the FDR.
It was noted that the 46 year old Captain, who had been PF for the accident flight, had accumulated almost all of his 3880 hours recorded flying experience on the accident aircraft type.
It was established that the aircraft had entered runway 18 from Taxiway 1 and had commenced a rolling take off simultaneously with a spot wind check from ATC of 270º / 28 knots. Subsequent examination of the recorded data from nearby anemometer showed that around the time of the take off attempt, the wind speed had varied between 14 and 27 knots at an average speed of 20 knots and the wind direction had varied between 261º and 291º and averaged 273º. AFM Limitations for the operation of the aircraft were found not to include a maximum cross wind component for take off but as the maximum demonstrated cross wind component for landing was 35 knots, it was noted that "operators often use 35 knots as a take off limit in case a return to the departure airport is required".
During the initial part of the take off roll, the ailerons were held into-wind by the First Officer and the Captain kept the rudder at neutral and maintained directional control using the Nose Wheel Steering (NWS). The first signs that directional control was being lost occurred as soon as the Captain had moved his left hand from the steering tiller to the control column which the First Officer estimated had occurred between 60 and 80 knots. FDR data indicated that the rudder had remained at neutral and the right aileron had remained into-wind throughout this time and tyre marks on the runway indicated that the aircraft had left the paved surface after it had travelled about 300 metres as it had been accelerating through 80 knots. As this happened, a right rudder command was made and maintained but it had no effect on the ground track of the aircraft which continued, yawing, in a straight line slowly diverging from the runway axis.
There was no call to reject the take off and at the point where both the FDR and CVR recordings terminated - at approximately 48 knots - the power levers were still "fully advanced". It was considered likely that this relatively early termination of recorder function had been the consequence of the loss of aircraft electrical power following impact damage to the avionics bay and its equipment.
The aircraft finally stopped almost 40 metres from the edge of the runway after an excursion distance of approximately 250 metres which had included crossing a disused runway. Once it had stopped, both pilots noticed that the remains of the propellers were still turning and whilst the Captain made a PA instructing passengers to remain seated, the First Officer operated the Fire Handles as a means to shut down both engines. This action caused the LP fuel cocks to close but would not have resulted in the immediate shutdown of the engines unless accompanied (as required by relevant QRH drill) by selection of the Condition Levers to the 'Fuel Off' position. The Captain stated that "he did not recall whether the condition levers were set to OFF during the evacuation".
No formal conclusion of the Investigation was recorded but in the absence of any evidence that the aircraft had not been airworthy, the following observations were made in respect of "Takeoff Technique":
- The technique for controlling the direction of the aircraft on the runway is to use rudder assisted by NWS at low speeds because the rudder has reduced effectiveness below 40 knots.
- Although the left pilot’s hand should remain on the NWS control until 80 knots (for directional control in case of a rejected take off), NWS should not normally be used above 60 knots.
- When rudder is used, the requirement for NWS to assist directional control will reduce progressively as speed increases above 40 knots and rudder effectiveness increases.
- It is likely that no assistance will be required by 60 knots and, therefore, there will be no step-change in NW directional effect when the pilot releases the steering control.
It was observed that during the accident take off roll, the rudder had remained at neutral throughout the time the aircraft had remained on the runway. In particular, it had remained so after the point around 40 knots where it would have become effective, with directional control having probably been maintained through the use of NWS alone. Had the rudder been applied as described in the Operations Manual, "there would have been a reduced NWS requirement at any given speed and therefore there would have been a reduced likelihood of a changed directional effect when the NWS control was released". It was, however, recognised that "the lack of data showing NWS commands meant that these considerations could not be verified".
In respect of the emergency evacuation, it was noted that no evacuation order had been given and the passengers had begun evacuation on theor own initiative. The Cabin Attendant reported that upon hearing a PA from the Captain for passengers to remain seated due to the still turning propellers, she had walked through the cabin to ensure passengers complied. Subsequently, she had then seen that "a passenger had opened the over-wing exit on the right side of the aircraft but, because the right propeller had now stopped and he was helping passengers to exit the aircraft, she (had) let him continue".
The Final Report was initially published on 8 October 2015 and this was then replaced by a slightly amended version on 13 October 2016. No Safety Recommendations were made.