On 24 December 2002, a SA 227 Metroliner III being operated by Danish freight and passenger charter operator Benair on a positioning flight from Aberdeen to Aalborg with just the two pilots on board crashed just after take off in marginal VMC at night following a loss of control. It collided with a car which caught fire and both aircraft and car were destroyed although only one person, one of the flight crew, sustained any injury, which was minor.
An Investigation was carried out by the UK AAIB. It was established that the First Officer had been PF and that the crew had believed that there had been a right hand engine failure shortly after takeoff from Runway 16 and in response, the aircraft commander had feathered that engine but had not raised the landing gear. After reaching approximately 100 feet aal, the aircraft had descended in a right turn with the PF stating that he "couldn't control the aircraft" and came to rest approximately 500 metres to the right of Runway 16 and approximately abeam the end of the runway having turned to the right by around 95° relative to the runway heading before impacting the ground. The plan and figures below showing the re-constructed aircraft trajectory is taken from the official report.
Radar and FDR Data, SW4 Accident, Aberdeen, 24 Dec., 2002; Source: AAIB Bulletin No: 6/2004 Ref: EW/C2002/12/03
It was noted that at about the same time as the PF had begun to experience an un-commanded turn to the right, a burning smell had been detected in association with an imbalance in engine EGT indication - the right hand engine being at 600 degrees C compared to the left at in excess of its allowable limit of 650 degrees with the associated fuel by-pass light. It was also noted that, although the crew had been unaware of any bird strike during take off, bird remains had been found on the departure runway shortly after the accident and there was evidence that the left engine (but not the right) had ingested birds. All the bird remains were found to the left of the runway centre line and included parts of two Herring Gulls, which have a typical weight of around 1 kg.
However, no evidence was found of a pre-impact failure in either engine, or any failure that could explain the loss of power on the right engine. The Investigation found that
“The detailed technical examination of the left engine and its propeller assembly revealed evidence of damage consistent with this unit delivering a high level of power at impact. The examination of the right engine and propeller revealed all damage to be consistent with a low, or no, power condition at impact, consistent with either a genuine loss of power or as a consequence of the commander pulling the 'Stop and Feather' control immediately before impact. However, an exhaustive examination of the right engine revealed no evidence of anything that could have caused a failure. Therefore, the items found on the runway were not considered to have been causal or contributory factors in the accident.”
It was additionally established during the investigation that the Cockpit Voice Recorder (CVR) had been inoperative and that the flight had been operated contrary to JAR-OPS requirements in respect of both the Flight Data Recorder (FDR) system and flight crew training.
It was considered that:
“although an engine failure during takeoff after V1 is a serious emergency, the aircraft was at a relatively light weight and, even with an such a failure, the crew should have been able to fly (the aircraft) safely away (from the ground). However, if other factors had been involved, the margins for safe flight would have become more critical.” These other possible factors were identified as:
- incorrect operation of the NTS system and/or a failure of the feathering system on the right engine,
- a concurrent power loss from the left engine,
- the flight crew not handling the emergency effectively.
It was also concluded that
- the aircraft commander's action in retarding the left power lever had led to a more significant reduction of power than any transient effect attributable to the bird strike.
- the amount of left rudder input by PF was likely to have been somewhat less than full deflection
- the landing gear had not been raised because a positive climb was not seen – yet FDR data showed that the maximum achieved airspeed had been 19 knots higher than the applicable V2 speed and that these last three factors “may have combined to reduce the climb capability of the aircraft to zero” and were all a result of “inappropriate crew actions”.
One Safety Recommendation was made to the JAA as a result of the Investigation:
- that the Joint Airworthiness Authorities should ensure that accident flight data recording systems fitted to aircraft that are required to be fitted with a Flight Data Recorder under the terms of JAR-OPS sample and record normal acceleration data at a rate of no less than eight times per second. (2004-31)
The Final Report of the Investigation was published on 3 June 2004 and may be seen in full at SKYbrary bookshelf: AAIB Bulletin No: 6/2004 Ref: EW/C2002/12/03