Australia’s air safety investigator says a simple typo, involving a pilot entering a ‘2’ when he meant to press a ‘3’ on a keypad was the most likely trigger for an incident in 2009 when an Emirates A345 crashed through the fence while trying to take off at Melbourne Airport and dropped down a slope toward a suburb before finally gaining enough lift to avoid a crash.
Its final report into the incident, which happened to a late night departure to Dubai on 20 March 2009, focuses on the international regulatory effort that has been underway since the accident to prevent what the ATSB describes as a widespread and serious issue for all airlines and all types of modern jets using electronic data entry to calculate takeoff performance.
But for the 275 people on board the Emirates flight EK407, and residents in Keilor Park, which was in its path after it burst out of the airport without becoming airborne, this was a near miraculous escape.
When the flight was being planned before departure, a pilot correctly calculated the take off weight of the jet as 362.9 tonnes, but entered 262.9 tonnes into the computer used to manage the take-off using the reduced thrust or ‘flexi’ settings that are commonly adopted by airlines to reduce both fuel consumption and engine wear and tear.
It wasn’t until 60 seconds after the take off roll began, and with the jet failing to rotate as anticipated, just one second away from reaching the red runway end lights, that the captain, who had been supervising the first officer who was making the takeoff, slammed the throttles into the maximum thrust TOGA (take off/ go around) detent, as shown in this extract from the flights performance as it roared down runway 16, which is 3657 metres long.
11.30.49 Brakes released, 3540 metres to go.
11.31.55 Rotation starts, 886 metres.
11.31.57 Nose gear uncompressed (off ground), 727 metres.
11.32.03 First of three tail strikes, captain orders full power TOGA, 229 metres left.
11.32.03 Levers moved to TOGA detent, 0 metres.
11.32.07 Main wheels uncompressed, 115 metres beyond end of runway.
11.32.09 Positive climb begins, 292 metres past end of runway (where the ground falls gradually below airport height).
11.32.46 Landing gear retracted (possibly near houses).
The final report clarifies that the captain applied full throttle, rather than call for it, as shown in the table above taken from the preliminary report.
The jet disappeared from view from the Melbourne control tower on its southerly departure after breaking through the airport’s perimeter with its wheels still firmly in contact with the ground, only for its lights to reappear as it climbed up from the hollow into which it had fallen before the engines could deliver enough power to pull away from the houses rushing up to meet it in the darkness.
EK407 was doing not less than 290 kmh when it broke out of the airport, briefly dropping down sloping grassy terrain beyond its southern boundary after ripping out part of an instrument landing system antenna array, and striking its tail on first on the tarmac and then open land.
Once the jet had enough velocity and power to climb it was taken out over Port Phillip Bay to dump fuel at an altitude of 7000 feet before making an emergency yet still overweight landing at the airport 64 minutes later.
The damage to the rear section of the jet was so extensive that it was given lengthy initial repairs rendering it flyable to Dubai, where it began a series of repairs that led to the complete replacement of the rear section of the fuselage and tail.
The ATSB finds that fatigue was not a factor in the entering of the incorrect data into the flight planning software, nor in the crew’s performance during the emergency.
It finds that some of the operational procedures set down by Emirates to prevent erroneous data being entered were not carried out by the pilots, but who were distracted by other matters while they were in the cockpit preparing for departure.
Emirates had already identified a need for ‘distraction management’ in the cockpit of jets being prepared for departure several months prior to the Melbourne accident, and was preparing a training course to deal with it but which had not been implemented when it happened.
The ATSB report says that Emirates late this year was working with a major avionics company to develop a take off monitoring tool that would alert pilots to discrepancies between actual and required take-off performance in its jets well before the point at which the four pilots in the cockpit of EK407 (including two relief pilots) noticed that they were in trouble.
It also reports that the US Federal Aviation Administration position in relation to erroneous take off data being used in airliners emphasises improved airmanship, while its European counterparts favor additional software or instrumentation, which the Americans regard as adding complexity to systems that may not prove as effective as better piloting.
Beyond the polite words in the ATSB report, it sounds like yet another disagreement between the US and European safety regulators as to how technology should be used to make flying safer.