August 03, 2005

Air Crashes and Data

In the following post, I speak of aircraft accidents only. This does not include terrorist attacks, bombs, people flying planes into buildings. This is strictly 'accidents'.

Today when I was getting my car worked on, the news was playing the aircraft accident that occurred in Canada yesterday with Air France. Wow. Sucks to be them. And to think they all got out… absolutely amazing.

As I was watching this, while reading… so my attention was not totally transfixed to the tube, they had this quote, “The NTSB says that 96% of all passengers survive an aircraft accident”. I wish I could find the exact verbiage as that’s REALLY important. Statistics can be twisted.

First thought when I read that was, “Yeah. Tell that to the folks that wandered out of that Iowa cornfield when that fan disk liberated.” 61% survived, although, I will say that any survivors of that accident are still viewed by me as a miracle in itself, even though I know why they survived. (An interesting read on that is HERE. It has the transcript and is a talk given by the pilot.)

But then I started thinking. What does the NTSB consider an accident? Their definition and mine are probably distinctly different. I’m sure that my definition is a subset of theirs, but theirs encompasses incidents I would NEVER contemplate. A hard landing resulting in bent landing gear isn’t on my scope, folks. Falling 30K out of the air and a big damn ball of fire is.

Of course airline crashes are going to make the news… and as I’ve said in the past, I don’t like flying. Loathe it. I become extraordinarily religious when I fly. But from what I understand, what made the Sioux City crash such a big deal to the public… is it was the first crash when there was a television crew on scene. There were television crews there through the entire event. What made it a big deal for those of us in the aerospace business was for engineering reasons; it had a serious effect on maintenance from an engine standpoint.

Anyway, I did some research. Of course I could find nothing on-line that showed the NTSB’s claim, which I don’t doubt. I don’t. I just want to know what they consider an accident. The best I could find was THIS… remarks from a symposium in 1999. I contend that he has some excellent points, although I wonder what their definition of 'substantial aircraft damage' is.

No offense, but a plane SITTING on a tarmac, at idle, not even making it down the runway, liberates a turbine blade, cutting through the fuselage and killing a couple people… that’s not an aircraft accident in my book. That's an aircraft incident. I’m sorry. It is an exceedingly unfortunate event that killed some people.

See... the number of people that rogue blade could kill was directly proportionate to the number of people sitting within its path as it tore through the fuselage. The plane wasn't moving. So you can't jack up a survivability rate with an incident when the survivors lives were never in jeopardy.

Accidents occur at take off and landings… they happen when something malfunctions in the air. I am sure my stats of survivability with only those accidents would be different. Once that plane is at full throttle barreling down the runway, the chances of living through something has taken a drastic plunge.

So I give them their 94% or 96% survivability, but it provides me no comfort… as their defintions for pulling data are different than mine... and it really really sucks to be that 4-6% that died.

Posted by Boudicca at August 3, 2005 09:40 PM | TrackBack

Are you going to do stats on car accidents now?

Posted by: vw bug at August 4, 2005 06:29 AM

Perhaps you heard them wrong, or the quote was read wrong -- perhaps they meant 94-96% do NOT survive. That sounds a lot more accurate to me.

Posted by: Ogre at August 4, 2005 07:48 AM

VW- I'll leave that to Machelle. I only know aviation. Being in that one group for all those years did not help my flying issues...

Ogre- Nope. It actually flashed on the screen. I read it. Amazing. Their definitions and mine are just oh so different.

Posted by: Bou at August 4, 2005 08:46 AM

Since you work in the field and are STILL scared, maybe I should rethink my views, but I've always thought the scariest thing about flying was the drive up Interstate 95 to the airport.

Posted by: George at August 4, 2005 08:52 AM

Actually statistically it is... the drive on I-95 is more dangerous. But if you're in a moving car and get in an accident, I think your survivability rate is much higher.

As for me? I just worked in the wrong group for a few years. Don't get me wrong, I frickin' LOVED that group, but sometimes sifting through all the bad stuff will skew your perceptions. Completely illogical, but there it is. One of my co-workers had a shrine to Amtrak in his cube...

Posted by: Bou at August 4, 2005 08:56 AM

Sometimes knowing "too much" can be a bummer.

Think of the poor guy who worked in sausage-making for 20 years going to a cookout and being served brats.

Or some poor doof who spent 20 years developing the chops to actually make music being forced to listen to top 40 manufactured, off-key (tone deaf), pseudo-music crap.

Or a guy who worked i n network news for 20-30 years who "got religion" and rediscovered his ethics and morals being compelled to listen to Dan Blather.

Knowing "too much" can just flat out ruin some experiences.

But sometimes the alternative (knowing too little) can have even more serious consequences.

Keep taking the train, Bou.

Posted by: David at August 4, 2005 12:31 PM

Oy! Thanks. My husband just told me that we had enough miles for the whole family to fly "somewhere". I can't tell you the anxiety I felt over this simple statement. I can't tell you the anxiety I suffer everytime he boards a plane for business. Don't get me wrong, I used to love flying and did it frequently. I've only flown twice since having my kids and not at all since 9/11. I've gotten really nervous about it. My husband wants to put my boys on a plane to Georgia from Texas to visit his brother for a week or so and I flat refuse. I won't do it. If that plane went down with all three of my babies I'd die as well. I feel better that at least if we were all together we'd die together. How messed up is that?

From my observation, when a plane has an "accident" it is more amazing when there are survivors, than not. Seems to me that when a plane plunges from the sky - everyone on board is pretty much screwed.

Posted by: Momotrips at August 4, 2005 01:11 PM

Think the difference is what you call it: Incident Vs. Accident. Statistically, flying is the safest mode of transport. Most things that can and do happen (incidents) are very survivable. It is when things go catastrophically wrong that it gets very bad. As a pilot, I trained for incidents and accidents, with the caveat that you were SOL on some things. Very bad things in my book were engine failure(s) on takeoff and major structural failures in flight -- not much you can do in either case. Engine(s) fail in flight or landing -- not such a big deal as I can and will land about anywhere, and the higher I am the more options I have. No matter what, though, it still sucks to be that 4-6 percent.

Posted by: Laughing Wolf at August 4, 2005 01:53 PM

Momotrips- "Seems to me that when a plane plunges from the sky - everyone on board is pretty much screwed." That sounds like something I would say! And I have this thing too about my family and flying. If we're going down, we're all going down.

David- yeah, I like the trains. A lot. ;-)

LW- That's the thing with statistics. People say, "Hey, only 1 person died in that hurricane!" yeah, that's great... unless you're that one damn person. Suddenly that 'only one person' is a really big frickin' deal!

Posted by: Bou at August 4, 2005 02:03 PM

You pays your money and you takes your chances...

If offered a ride on the next space shuttle mission, with no additional fixes after the current flight, would I go?

Hell, yes!

Statistics bite us all in the ass eventually, so I'm going to *live* until my behind is in the mouth of Hell.

And even then, I'll still do the "young and foolish" thing...


Posted by: Jack at August 4, 2005 02:54 PM

I have no desire to go into space. No thanks. I'll stay firmly on the ground! I don't live on the edge much.

I have a very healthy grasp on my imminent mortality...

Posted by: Bou at August 4, 2005 10:53 PM

Found this report online from the NTSB...

I'll warn you now... it's excessively repetitive even though it's not very long. Save yourself the aggravation and skip to the chapter on Aviation Accident Survivability. (I don't know who proofed this thing but I swear I've read the same set of sentences at least 4 times at various places) Anyhow... for those not interested in skimming through it - an "accident" is defined as follows

"an occurrence associated with the operation of an aircraft... in which any person suffers death or serious injury, or in which the aircraft receives substantial damage".

They are quite right when they say that the news idiots just don't report accidents that are non-fatal or rather - there may be an offhand report in a very localized area. Rather like the reportage on a car accident that totals a car, but the driver is only somewhat injured vs an accident where all involved die at the scene...

The thing about most people and aircraft is that they know they don't have any control - it's very difficult to accept that someone else is doing the driving and if bad stuff starts to happen, there's nothing you can personally do to make it better.

As for me - I love to fly. I figure if I go in an air crash - well at least it's better than a lot of other ways I could die. *grin* (wonder if this makes me a fatalist?)

Posted by: Teresa at August 5, 2005 01:22 AM

Teresa- It sounds like the same file I have linked, except in PDF form. All the same information you quoted. That's why I was saying, "What's their definition of substantial damage?"

They just view things differently than everyone else. But it is the nature of the business.

Posted by: Bou at August 5, 2005 08:37 AM

"The thing about most people and aircraft is that they know they don't have any control - it's very difficult to accept that someone else is doing the driving and if bad stuff starts to happen, there's nothing you can personally do to make it better"

Teresa, you just hit my nail on the head. That is the very reason I do not like to fly. I have no control at all over whatever happens. I hate having no control in any situation.

There was only one time recently that I didn't mind flying. What happened was the Crew (pilot and other pilot) sat next to me in the waiting area waiting for the plane to come in. It was delayed and we struck up a conversation. We talked for an hour. By the time we were flying I felt like I knew them and felt comfortable with their abilities.

I don't like putting my life in the hands of someone I have never met in my life.

Posted by: Machelle at August 5, 2005 08:46 AM

Bou - on page 9 of the pdf file - the footnote number 10 defines substantial damage as the following...

...damage or failure that adversely affects the structural strength, performance, or flight characteristics of the aircraft, and would normally require major repair or replacement of the affected equipment.

I'm thinking they don't mean something like a blown tire. Although that would affect the ability of the aircraft to take off and land - that isn't a stuctural issue. OTOH I could be wrong... wonder if anyone from the NTSB reads you. *grin*

Posted by: Teresa at August 5, 2005 04:17 PM

Yeah, I'm only very familiar with the military definitions... Class A mishap and all. Either way, my data pool would be different than theirs. A turbine blade liberating on the tarmac while engines at idle will cause that kind of damage. (I'm citing a true incident.) It causes substantial damage when a blade comes through the engine compartment, cutting through the fuselage. All those folks not in the path would be considered survivors. I wouldn't consider them at all.

And yes, a blown tire could concievably do it. If you have a hard landing that damages the landing gear, scraping the bottom of the airframe, that fits their billet for substantial damage. It's secondary damage would fit their "damage or failure that adversely affects the structural strength".

Posted by: Bou at August 5, 2005 04:42 PM

NTSB Part 830 (available at is what you need. Damage to landing gear is not an accident, but I think your example of the turbine blade would be an accident - passengers have embarked for flight, and something happens which causes substantial damage. By NTSB 830, that's an accident. Now, if the damage was confined to the engine, it wouldn't be an accident.

Posted by: Bob at August 10, 2005 12:52 PM