November 2009 Archives

The second "interim factual" report to the Qantas A330-300 that dove 1000 feet in minutes has just been released [opens as PDF].

Airbus is modifying the flight control software in A330s and A340s. An interim modification will be made to all Qantas A330s by the end of this month. A standard modification for all A330s and A340s will be certified next year and then retrofitted to all A330s and A340s.

Big news item so far:

In addition to the initial procedures-based safety action taken by the aircraft manufacturer in response to this accident, Airbus is modifying the flight control primary computer (FCPC or PRIM) software used in the A330/A340 fleets to prevent any future similar problems leading to an uncommanded pitch-down event. An interim modification to the FCPC software standard is being installed in the operator's fleet, and the installation is expected to be completed by the end of November 2009. A later FCPC software standard to improve the treatment of all ADIRU parameters will be certified in mid to late 2010, and will then be retrofitted to the world-wide fleet of A330/A340 aircraft.

(Emphasis added.)

Update:

The problem
One of the three ADIRU units is at fault, as the ATSB reported in its first report. But the big question of what caused the fault (i.e. erroneous data resulting in the aircraft plunging) still remains unknown.

1:41 PM--Mainstream media reports are now trickling out, and one article is putting possible blame on "cosmic rays". While the report does mention cosmic and solar radiation have the potential to interfere with systems, the official language is "The investigation team is evaluating the relevance, if any". So don't go wrapping yourself up in aluminum/tin foil quite yet.

What the ATSB is focusing on
Since this incident is caused by software gone astray, the ATSB will "examine various aspects of the flight control primary computer (FCPC or PRIM) software development cycle including design, hazard analysis, testing and certification." The conclusion will be presented in its final report, due out in 2Q 2010.

What the ATSB has done since the first report
The ATSB has disassembled, examined, and tested the faulty ADIRU from the Qantas A330-300 that operated QF72. This did not yield any information related to the accident. Additional testing may be carried out.

The ATSB installed a different ADIRU on the accident aircraft, VH-QPA, and performed a number of checks during a 11 hour test flight (with no revenue passengers) around Australia in May 2009. There were no abnormalities.

Relation to other incidents

• Jetstar A330-200 incident (February 2008)
  

At the time of releasing its first report into this accident, the ATSB was focusing on similar ADIRU failures (but which didn't cause any dives or other "upsets"). One incident involved a Jetstar A330-200, whose ADIRU were made by the same manufacturer of the Qantas A330-300s,  but the ATSB says it now believes the Jetstar incident is unrelated to the Qantas incidents. (There were three incidents on two A330-300s.)

• Air France A330-200, AF447 (June 2009)

While the ATSB noted the French investigation is ongoing and the reasons for the accident have not yet been determined, the ATSB also pointed out a number of differences between the Air France and Qantas incidents. The ADIRU units and pitot probes on the Air France and Qantas aircraft were each made by different manufacturers. Cockpit messages on Air France and Qantas showed a different sequence and pattern of events.

What next?
The ATSB says it is going to continue focusing on 1) why the ADIRU provided erroneous data, and 2) the role of flight control software.

A final report is due out in the second quarter of 2010.

Following Tuesday's release of the Australian Transport Safety Bureau's (ATSB) tests on oxygen bottles, tomorrow the ATSB will release its final second interim report on the Qantas A330 that dived 1000 feet in a matter of minutes over Western Australia.

Let's have a recap of the event and the first interim report (opens as PDF).

Basics

• Flight: Qantas 72 from Singapore to Perth (2420 miles), 7 October 2008
• Aircraft: A330-300 VH-QPA
• On board: 303 passengers, 9 cabin crew, 3 flight crew
• Injuries: 11 passengers and 1 flight attendant sustained serious injuries, others had less serious injuries. Most occurred as the individuals were standing or not wearing seat belts. No fatalities.
  
• There was no structural damage, but the centre and rear cabin sustained damage. See photos:
  

Timeline (in local time)

• 9:32 am Aircraft departs Singapore
• 12:40:24 pm At cruise altitude of 37,000 feet, autopilot disconnects, cockpit gives system failure notifications
• 12:42:27 Flight crew evaluating situation when aircraft pitches nose-down at a maximum angle of 8.4 degrees, descends 650 feet. Crew brings aircraft back to 37,000 feet, deals with failure messages
  
• 12:45:08 Aircraft commences second commanded plunge, this time at a maximum angle of 3.5 degrees, and descends 400 feet (approx 160 seconds after first plunge)
  
• 12:49 Flight crew declare PAN urgency, begin diversion
• 12:54 Flight crew learn from cabin crew of serious injuries, declare mayday
• 13:50 Aircraft lands at Learmonth

Aircraft's flight path (ATSB):

What's at fault?
The preliminary report ("interim factual") identified two concerns

1. One of the three air data inertial reference units (ADIRUs) "started providing erroneous data (spikes) on many parameters to other aircraft systems". The other two ADRIUs functioned normally.
2. The second concern is a bit more technical: "Spikes in angle of attack data were not filtered by the flight control computers, and the computers subsequently commanded the pitch-down movements".

More simply, as our Kieran Daly on the safety beat summarised the accident, "the crew was trying to control a fly-by-wire aircraft which didn't 'understand' what it [the aircraft] was doing or what was happening to it".


Significance
At the time, the ATSB was investigating two similar incidents involving "anomalous ADIRU behaviour" but which didn't involve an "in-flight upset".

1. One incident involved a Malaysia Airlines B777-200 240 km from Perth. The ADIRU was from a different manufacturer. (Read the report in PDF here.)
2. A second incident involved the same aircraft, VH-QPA, as the one concerned in this aircraft. The incident, in September 2006, involved the exact same ADIRU. After the  flight, the unit was inspected and re-set.

Additionally, on a December 2008 flight from Perth to Singapore, a different Qantas A330-300 received a fault message regarding the ADIRU. While it was a different ADIRU from the other incidents, it was in the same position (the #1 ADIRU).

ATSB map of the three Qantas A330 incidents (not included is the Malaysian 777 incident):
 A greater incident would occur three months, with the complete loss of Air France flight 447. While that incident focused on faulty pitot tubes, an ACARS message reported a fault in an ADIRU.

Report's Influence
The report ruled out the dive as being a result of simple turbulence and advanced military operations that interfered with the aircraft's systems. It also ruled out the rumour a laptop or wireless headphones (media speculation couldn't settle on which one) was at fault.

The algorithms used to calculate data relating to the ADIRU are also featured on the A340. EASA and the FAA issued air worthiness directives. Of the 900 A330 and A340 aircraft in operation at the time, 397 had the same ADIRU model as the one involved in the accident and other incidents.

For its investigation, Airbus searched for similar incidents within a designated time frame. It could only identify four: the Malaysian 777, the previous incident on VH-QPA, the incident in December 2008 on another Qantas A330-300, and finally an incident on a Jetstar A330-200 (VH-EBC) from Sydney to Ho Chi Minh.

What's going on with A330s in the skies around Australia?

Check back here tomorrow for an update on the second report.

Don't take life too seriously.

That is the unofficial mantra Down Under. It helps to explain the irreverence in Australian humour, such as why a person with red hair is called a "bluey"--a joke the red-themed Virgin group used to name its Australian start-up: Virgin Blue.

Life may be less serious. Life may also be a bit more remote: Australia and New Zealand are 1500 miles apart and the next closest (major) international city is some 3400 miles away.

But it's worth keeping an eye on aviation in this region, which this new Flightglobal blog aims to do. So what's so special?

As the world plans to attempt to tackle climate change next month in Copenhagen, Air New Zealand is at the forefront of improving aviation's impact on the environment. It flew the first sustainable biofuel flight, aims to have 10% of its 2013 fuel needs met by biofuels, and this week in Dubai announced it would be the launch customer for winglets ("sharklets") on its new A320s.

In Australia, Qantas had it good for so long. Then Virgin Blue entered in 2000, creating a low-cost domestic scene, forcing Qantas to respond with its "two brand" approach: Qantas for full-service and Jetstar for low-cost service.

The golden rule was Jetstar would never compete with Qantas, but now as Tiger Airways (below) is encroaching, the two brand approach is being put to the test domestically. Internationally, Jetstar was supposed to have 787s by now, flying them to Europe. 787s have been pushed back to 2013 and now the carrier is looking at using A330s, but can they take the payload? Has low-cost, long-haul operator AirAsia X taken the market?

Back to Qantas, it is the world's second oldest airline and is deeply held by Australians, even when oxygen bottles explode mid-flight, an aircraft loses a thousand feet of altitude in seconds, and pilots forget to put down the landing gear.

Virgin Blue launched in 2000 as a low-cost carrier but soon moved up-market as the world's first "new world carrier", an appellation it has had to defend as something other than a moniker. The carrier now operates more domestic flights a day than Qantas, and is making an impact on the international scene.

Virgin Blue launched V Australia last February, arguably the worst time ever, but had little choice. Soon it is starting services to Phuket, Fiji, and Johannesburg. The grapevine screams (and the carrier quasi-denies) V Aus is looking at ordering 787s and 777-200LRs, which it would use on non-stops from Sydney to New York and Perth to London, creating the world's first and third longest flights, respectively. Maybe it's not happening as soon as reports made it out to be, but the rumour is it's on the eventual agenda.

The Australian and New Zealand regulatory regime is unique in that foreign companies can launch domestic airlines with no majority-owned regulatory constraints, such as those that plagued Virgin America. That's what enabled Virgin Blue to initially launch (it's now Australian majority-owned) and ditto for Tiger Airways Australia, a sister airline to Tiger Singapore. Tiger proclaims itself as Australia's real low-cost airline and in the past six months has taken on a major expansion plan, breaking out of its Melbourne and Adelaide hubs.

That's just the beginning. There are smaller carriers, regional carriers, modern air traffic management procedures, and lots of defence projects in the swathes of Australia's desert.

Stay tuned.