A dark and windy night

July 1, 2009

Miles O’Brien is a pilot, airplane owner and freelance journalist who lives in Manhattan. His blog is located at www.milesobrien.com. The opinions expressed are his own.

A lot of travelers boarding an Airbus today might be thinking twice. After all, yet another Bus is at the bottom of yet another ocean – and another 153 souls have gone west.

Could the European airliners be latter-day versions of the DC-10? That is, a flawed design and thus a relatively dangerous way to fly?

For the entire Airbus airliner fleet (more than 5400 are in service globally), the numbers do not support that conclusion.

In July 2008, Airbus’ bitter rival Boeing released a “Statistical Summary of Commercial Jet Airplane Accidents” from the dawn of the jet age in 1959 through 2007.

At the time of the study, the A330 still had a flawless record: no fatal accidents in the course of a million departures. A month ago, Air France 447 changed that record, but the airliner remains very safe statistically.

Over the years the Airbus A300 has had three crashes that caused deaths. That equates to a rate of .47 airplanes lost per million departures. The A320 series has had eight fatal crashes – or .23 hulls per million departures. And the A340 has never had a fatal crash.

A Yemenia airlines Airbus 310-300 registered under the number 70-ADJ taxis on the tarmac of Charles De Gaulle International Airport in Paris in this July 27, 2002 file photo. REUTERS/Thomas Noack (FRANCE DISASTER TRANSPORT IMAGES OF THE DAY)The record is not as good for the A310, the model of airplane that plunged into the sea trying to land at Moroni, the capital of the Comoros Islands. It has crashed and killed people eight times now. That equates to a fatal accident rate of 1.42 airplanes for every million departures.

(The actual plane that crashed is pictured shown here in a photo taken in 2002.)

The infamous and much maligned DC-10 crashed with fatalities a dozen times, for a rate of 1.36 fatal crashes per million departures. Pretty much a dead heat (if you will pardon the expression).

It is worth noting that these fatal accident rates have come a long way. The early jet airliners – the 707 and DC-8 – logged fatal accident rates of 4.21 and 4.03 per million departures respectively.

But take a look at the accident reports for the A310 — there are two common threads. First, they are all attributed to pilot error: Trying to land in a thunderstorm, botched use of thrust reversers on rollout, improper stall recovery, spatial disorientation on a dark, stormy night, a botched missed approach, and the most infamous of all, the captain who allowed his son to take the controls, leading to a stall and spin.

The second is the airlines were all flagged in third world/emerging nations [Maybe the Russians might quibble with that characterization, but over the years Aeroflot has logged a third-world quality record.]

This is why you are hearing so much talk about the so called blacklist of airlines that are banned from flying to Europe or the U.S.

Airlines have to be pretty sloppy (and scary) to get on this roster. It means they lack:

• the regulations to properly certify airplanes

• the technical expertise and resources to oversee them

• adequately trained technical personnel

• adequate inspectors to insure they comply with minimum international standards

• and insufficient record keeping to document what they are doing (or not).

Yemenia Airlines is not on the European blacklist, now 194 airlines long. But the crashed 19-year-old/17,300 cycle airplane (7O-ADJ) apparently was, at least in France, where it was banned in 2007 because inspectors there found long list of squawks.

So why so many pilot error crashes by crews flying the A310 for third-world airlines? Is it shoddy training? Is it simply that the A310 is a cheap, widely used aircraft for thinly endowed airlines? Is it the flying environment in the countries where these planes fly, with fewer, less sophisticated navigational aids and less air traffic control coverage and expertise?

Could the highly automated Airbus design be ill-suited for these crews/ airlines/airports? Or has it saved untold lives by preventing accidents? These are hard questions to answer.

Yemen Airways flight crew alight from a company bus at Sanaa International Airport July 1, 2009. The flight recorder from the Yemenia-operated Airbus A310-300 that crashed into the Indian Ocean on Tuesday with 153 people on board has been located, the airline's head said on Wednesday. REUTERS/Khaled Abdullah (YEMEN DISASTER TRANSPORT BUSINESS)Unlike Air France 447, we should know the answer to this riddle fairly soon, as searchers have already found the black boxes (left).

But the man in charge of the airline claims he knows what happened.

“We never had problems with the plane,” Yemenia Chairman Abdulkalek Saleh Al-Kadi told Bloomberg. “It was purely weather.”

What about the weather? Here is the weather picture (in pilot parlance, a METAR) for MORONI/Prince SAID IBRAHIM (FMCH) airport:

FMCH 292300Z 21025G35KT 9999 FEW020 25/16 Q1017 TEMPO 18015G30KT

Translated – it means the wind was coming out of the southwest (210 degrees) at 25 knots (28 mph) gusting to 35 knots (40 mph). There were a few clouds 2,000 feet. So it was windy and the sky was nearly clear albeit totally dark when the crash occurred just before 2 AM local time, and moonset that night was 12:23 AM.

With that in mind, let’s try to imagine ourselves on that Yemenia flight deck. The Moroni airport has one runway that allows planes to land either toward the northeast (20 degrees) or the southwest (200 degrees). Airplanes nearly always land into the wind, especially when it is blowing as strong as it was at FMCH that night.

But there is only one precision instrument approach to the airport – and it is for the runway that would have forced them to land with a strong tailwind. So the crew was forced to fly a visual approach to runway 20 on a dark night over water – approaching an island that probably does not have many lights blazing at that hour.

PAPI Lights - Wikipedia

To add to the challenge, runway 20 does not have a Precision Approach Path Indicator (PAPI) (left). This is an array of focused light beams that sit beside a runway and give a pilot a visual indication of where his craft is relative to the ideal glide path. A four light PAPI – as you see here will show the pilot two red and two white lights when he/she is at the correct altitude for a safe approach. More red – and you are too low…more white and you are too high. It is truly pilot-proof.

But without those lights on that dark night over the water, the crew would have had a hard time judging how close they were to the ground (or the surface of the sea). It is called “spatial disorientation” and it kills a lot of pilots and passengers (including John F. Kennedy, Jr., his wife and sister-in-law).

They apparently tried to land once – but aborted the approach – turning around in a “black hole” – itself a perilous maneuver – especially for a crew that would be a bit rattled and distracted by their predicament – and were, no doubt, dog tired after a long day of flying.

It is the perfect recipe for losing focus on your gauges – and forgetting which way is up – and how far is down.

(This commentary was corrected to change the landing direction for planes at Moroni airport to northeast (20 degrees) or the southwest (200 degrees).)


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/

The article is simply saying that all fingers should be pointing away from Airbus. This is true considering the statistical track record of its airplanes. Airbus, especially after the recent two crashes, should be so grateful for such objective, non-feverish article.
Despite the great concern about the safety of the international civil aviation, there are still some airports, airplanes, and operators acting against the trend. Some more concrete regulations/ actions should be in place to ensure the safety of the innocent airline passengers.

It is true that some Russian airlines have some operational problems. However, I think that talking about Aeroflot was totally out of context in the article for the following reasons:
• The aircraft is non-Russian made.
• The aircraft is operated by Yemenia.
• Departure and arrival airports of the flight are non-Russian.
• No Russian citizens were reported among the crash victims so far.
• The updated list of banned airlines by EU does not include Aeroflot (http://bit.ly/xs3ps).
• Aeroflot has a daily flight to JFK which simply means that it is not banned by US aviation authority.
• The same source quoted in the article (Aviation safety net) is also saying that USA has the highest rate of civil casualties (http://bit.ly/QJFA2).
• Aeroflot has achieved TRUEngine status for its CFM56-5B engines which are powering its Airbus family aircraft (http://bit.ly/13smLI).

Posted by Ahmed Sultan, ITC | Report as abusive

Miles I’m very surprised you include the following line at the start of your article.

“Could the European airliners be latter-day versions of the DC-10? That is, a flawed design and thus a relatively dangerous way to fly?”

Although you do indeed stress later on in the article that Airbus has a very good track record safety-wise. You should never have brought that up entirely.

As a pilot certified on the 737, currently active with a major EU airline, it is obvious to me why this happened.

The crew decided on a visual approach, at night, without any form of vertical or lateral guidance even without PAPI’s.

This course of action is reckless, dangerous and to my knowledge prohibited by all major EU airlines.

At the limit a circling to land procedure could have been attempted. However I am not familiar with the approaches to FMCH and so can not say if these are allowed.

To conclude, there is a VOR-DME station available at the start of runway 20. Unless proven otherwise both DME and altitude indications were available to the crew during the approach.

Ignoring that fact that the visual approach should never, ever, have been commended, There is still no excuse for both pilots failing to uphold the very basics of flying.

Posted by Albert Koughlin | Report as abusive

Lots of data here, but just data. I have flown into small airports at night, in cross winds, snow, and other hazards. I also have ridden in commercial planes in Alaska many times on some pretty nasty approaches into some pretty small airports. I was fortunate to have a pilot who believed his instruments and who had been in the same predicament many times. I think that when this all shakes out we will know the story. The Airbus with all its safety and automated systems can cause a pilot think that the plane is safer than it is. I would wonder how many landings these pilots had in tight situations with smaller aircraft.

Posted by f belz | Report as abusive

Seriously, how can such sloppy copy be allowed up on your otherwise reputable site? Is anyone copy-editing this?

Posted by N. Deutsch | Report as abusive

How did this piece of writing get published on the Reuters website in rough draft condition? Looks like the editor’s working on autopilot here :-)

Two observations: Ahmed Sultan (the writer of the first post) needs to take a paranoia management course; and it’s probably best to stick to flying (a) with first-world airlines (discount ones perhaps excluded) and (b) to first-world destinations (Air France flight 447 notwithstanding).

Posted by G Diss | Report as abusive

A long time ago, the BBC’s “Horizon” series (it’s screened as “Nova” in the US) made a documentary about these issues. At the time, the Airbus was still new and one had recently undershot the runway at the East Midlansd airport (I think) and nearly crashed onto a busy motorway.

Although the official cause of the crash was pilot error, a cockpit reconstruction revealed a (to a software guy) shockingly poor user interface for some of the pilot’s most critical instruments. I’d say that any place which consistently showed high levels of accidents attributed to pilot error should be looked at more closely from a UI point of view.

But then, I’d hope that the various air safety boards are already on to that one.

Posted by Ian Kemmish | Report as abusive

Just to clarify – the reason why Aeroflot was mentioned in the report is that the infamous episode of the pilot handing controls to his son which led to the crash occurred on the Aeroflot flight (http://en.wikipedia.org/wiki/Aeroflot#I ncidents_and_accidents).

Posted by nck7 | Report as abusive

Congratulations, interesting article.
I think it wise to hold opinions is very important to first have and then analyze the data. Take into account:

1)The report data 27,455 plane fatalities comes from Boeing:
http://www.boeing.com/news/techissues/pd f/statsum.pdf

2) NATO Secretary General George Robertson said that 1.2 million people die and 50 million are injured each year around the world due to traffic accidents. Thus, it is possible to estimate world car fatalities 1959 to 2009=1,200,000x 50years= 60,000,000

3)The aircraft fatalities is approximately 0.05% compared to deaths from car accidents= =(27,455 /60,000,000)x100 = 0,0457%

Posted by JGS | Report as abusive

Ian Kemmish
the British Midland plane that crashed was an 737-400!

Posted by nick | Report as abusive

Stay away from Airbus? Little harsh statement! Now I am trying to figure what did you wanted to say, but whatever you meant you are blaming only A310 (partialy) and un-trained pilots and airlines with poor maintenance record.

Posted by al | Report as abusive

very simple bring out a law that prohibits visual landing,exceptions only if all else fails,any pilot not following this, should be suspended for a month without pay,while they are flying they are responsible for the lives of all the passengers,when you purchase a ticket it is not for the reason to crash into the ocean or mountain or in a forest or a dinosaur,but to make sure every body arrives safely, including the pilots.

Posted by cyrus vanhelsing | Report as abusive

Miles,if that plane had RNAV the pilots could have built an approach with a glide slope(called VNAV).

Posted by Larry | Report as abusive

Very interesting post. I am also amused by the cross-through in the copy, which I have never before seen on a major news site.

Perhaps the airlines and the news organizations have something in common – having to act very quickly with too few staff and too little budget

Posted by Jim Fowler | Report as abusive

With very little experience but pilot with seaplane rating and helicopter time and solo I pronounce this interesting to the finale. You need more flying hours to really judge a matter like this

Posted by Ron Moss | Report as abusive

Well looking at the charts, (the ones I found are in french so I am possibly wrong) the visual approach to RWY 20 circles out over the ocean and rolls out to establish yourself basically on the on the centerline. Wind at the time was only 10 degrees off head-on.

But if it were me, I would stay a little to the right of center, over the ocean, because there are tall scary mountains to the left. Assuming good visibility, it seems pretty simple: or am I dead?

Posted by Tom | Report as abusive

When will the industry start to really address human factors? Seems to me humans have more to do with crashes than planes or cars or trains.

Posted by bill smith | Report as abusive

Flight 447:
(I’m not an expert but putting together what I’ve heard over alifetime)..

A plane stalls if it’s going to slowly to get lift.
In a stall, there is no control and it’s difficult to get the plane into dive to regain speed.

The pilot would try to get a nose-down attitude.

If it stalls, it tends to “pancake” – land on its belly.

It follows, if the plane is at (say) 8000 m, where the atmosphere is thin, providing little lift (which comes from air speed and air density), with narrow wings (to minimise drag, no flaps down), and its speed falls, it will stall, I believe.

If it’s in turbulent air, to regain control is more difficult.

If the plane is slowed because it’s in turbulent air, but slows to much (?iced pitot tubes), it will stall.

If it’s found to have hit the sea belly down or somewhat nose-down (as reported), it sounds like a stall.

It sounds like a stall.

Is this an airbus weakness, or what would always happen? I don’t know.
But it sounds like a stall. END

Posted by malc cochran | Report as abusive

The real problem I see with this page is the reporter mentions Airbus and DC and does not analyse Boeing safety.

Why not mention the crash rates of the best and worse Boeing planes?

Bottom-line is it’s another report with a bias against Airbus and Europe.

Posted by georges | Report as abusive

Good article.
I’m a white-knuckler on anything bigger than a single-engine private prop plane. There’s something about the feeling of sitting in a tube with tiny windows versus sitting in a cabin of 30% window area that gets me on edge.

Regarding Airbus, I recall the 320 had enough problems about it I’d check the type of plane prior to ticket, and if an Airbus I’d make other arrangements. I don’t recall exactly HOW the plane type was made available, but I do clearly recall switching to a different aircraft. My ex thought it ridiculous until she saw this recent wave of A3xx disasters.

In the current dismal economic climate, the airframe czars sure have their hands full with the Boeing vs Airbus game. I just hope it doesn’t get to anything more than a “Ford vs Chevy” thing. There’s been an increase in “crash survivability” info on the net, too.

The best result out of this Airbus awareness will hopefully be heightened quality of parts and control systems, which will improve both construction and maintenance.

And maybe they’ll finally get around to installing those massive parachutes to bring an aircraft to a gentle landing in case of impending disaster. Sacrificing a few seats to make room for such a system could save an entire plane’s worth of human cargo.

Posted by richie | Report as abusive

Very interesting assumption of what “COULD GO WRONG” under the circumstances, but an assumption nevertheless. Without the details from the flight and voice recorders, anybody’s guess of what actually happened, is as good as anybody’s else. Let’s wait and see the facts.

Posted by jeanmark | Report as abusive

i am a 100 000 miles a year passenger ww travel

for me NO AIRBUS for a while. period!

Posted by dr. hans ‘ROW’ roswell | Report as abusive

While taking off in an American Airlines Airbus from a rough runway airport in Puerto Rico, it was obvious looking down the aisle from the back of the plane that the whole length of the plane was twisting and torquing. The overhead bins were actually ‘snaking’ backing and forth in response to the undulating runway. This is not something you want to see in an airplane. No wonder they are finding ‘stress cracks’ in the carbon fiber sections. I avoid flying on an Airbus when possible.

Posted by Michael Echols | Report as abusive

Wasn’t Flight 447 meant to be flying over the weather?
I thought thunderstorms were well below 40,000 ft. Lightning doesn’t strike upwards, or does it?

Posted by dennis | Report as abusive