Cirrus... is it time for certification review?

And in the past year, the numbers have gotten worse.
Accidents and incidents (from theFAA and NTSB databases)
1998 2000 2001 2002 2003 2004 2005 2006
SR20 1 0 2 3 0 1 2 7
SR22 - - 2 2 3 8 12 15

We don't know if these are "worse" without knowing fleet size and hours
flown for those years.

Jose
--
"Never trust anything that can think for itself, if you can't see where
it keeps its brain." (chapter 10 of book 3 - Harry Potter).
for Email, make the obvious change in the address.

"john smith" wrote in message
...
In article ,
"Kyle Boatright" wrote:

Flying magazine (or AOPA?.. dunno) ran the numbers a year or so ago and
compared the accident rate between Cirrus and competitive models. I
don't
have a copy at hand, but there was a significant difference in accidents
with Cirrus having a much higher rate than the other A/C.

And in the past year, the numbers have gotten worse.
Accidents and incidents (from theFAA and NTSB databases)
1998 2000 2001 2002 2003 2004 2005 2006
SR20 1 0 2 3 0 1 2 7
SR22 - - 2 2 3 8 12 15

As someone else pointed out, you have to consider the number of aircraft in
service and, even better, estimate the fleet hours for the time period. The
article I mentioned attempted to do those things. A simple count of
accidents won't.

KB

Doing a little math:

And in the past year, the numbers have gotten worse.
Accidents and incidents (from theFAA and NTSB databases)
1998 2000 2001 2002 2003 2004 2005 2006
SR20 1 0 2 3 0 1 2 7
SR22 - - 2 2 3 8 12 15
TOTAL 1 0 4 5 3 9 14 22
rate (%): 50 0 2 1 .3 .6 .7 1
fleet size: 2 7 206 514 902 1491 1949 2323
SR22 fleet 121 383 687 1180 1560 1848
SR20 fleet 2 7 85 131 215 311 389 475

So, each year a bit less than one percent of the fleet bites it. The
rate seems to be increasing slightly in the last few years, but the
sketchiness of this data precludes a conclusion based on that.

To compare with the Cessna fleet (bearing in mind the errors in the year
data due to registrations), I'll just add the last five years of fleet
size, getting something like 125,000. Five years of accidents at a 3/4%
rate (the last five years of the Cirrus rate, eyeballing it) would imply
something like a thousand C-172 crashes.

So, were there "something like a thousand" C-172 crashes in the last
five years?

Jose

Fleet info source from 's post Oct 28, 1:10 pm,
summed for SR20 and SR22. I added the total fleet size (by
airworthiness date), figuring it was unlikely that the Cirrus fleet
would have accumulated many date errors yet due to sales.
--
"Never trust anything that can think for itself, if you can't see where
it keeps its brain." (chapter 10 of book 3 - Harry Potter).
for Email, make the obvious change in the address.

On Sun, 29 Oct 2006 03:34:05 GMT, Jose wrote:

So, each year a bit less than one percent of the fleet bites it. The
rate seems to be increasing slightly in the last few years, but the
sketchiness of this data precludes a conclusion based on that.

To compare with the Cessna fleet (bearing in mind the errors in the year
data due to registrations), I'll just add the last five years of fleet
size, getting something like 125,000. Five years of accidents at a 3/4%
rate (the last five years of the Cirrus rate, eyeballing it) would imply
something like a thousand C-172 crashes.

So, were there "something like a thousand" C-172 crashes in the last
five years?

From January 1st, 2002 to December 31st, 2004, the GA average fleet accident
rates were as follows:

Overall: 0.58%
Homebuilts: 0.80%
Rotorcraft: 1.63%
Robinson: 3.83%
Cessnas: 0.56%
Cessna 172: 0.62%
Piper: 0.47%
Piper Super Cubs: 1.02%
Beech: 0.45%
Beech 33, 35, & 36: 0.43%

To get the above results, the total number of accidents in the three-year period
were divided by the total aircraft of that type registered on 1 January 2005,
and the result divided by three to produce a yearly average.

Note that the Beech, Cessna, and Piper figures may be artificially low, due to
old aircraft that are still on the registry but not actively flying. Aircraft
can be abandoned or even scrapped without telling the FAA, hence they remain on
the register. The FAA is currently working on weeding out these old
registrations.

I haven't run the fleet accident rates for the Cirrus....guess maybe I'll have
to take a look. If, as you say, the accident rate is about 0.75%, that's in the
ballpark of the 172.

Ron Wanttaja

On Sun, 29 Oct 2006 01:37:47 -0700, I wrote:

I haven't run the fleet accident rates for the Cirrus....guess maybe I'll have
to take a look. If, as you say, the accident rate is about 0.75%, that's in the
ballpark of the 172.

Just completed a cursory look at the accident data.

A summary on the process: I previously downloaded the NTSB accident report
databases for the years 2002, 2003, and 2004 and developed various database
queries to allow analysis of accident causes. I've also downloaded the FAA
registration database every January for the past ~7 years. I have developed
various tools to allow comparison of accident rates.

Rather than compare the Cirrus accident rate to those of the overall fleet of
Cessna 172s, I have compared it only to the accidents involving 172s
manufactured since 1994 (yes, production started a few year after this point).
I have included any Cessna 172 listed with a post-1994 manufacture date, PLUS
all Cessna 172R and 172S models since the year of manufacture is not always
available).

For both aircraft, I have included only those accidents that occurred in the
United States.

After looking at my tools again, I see my earlier posting mis-stated my method
of determining the total fleet size of a particular aircraft. I had stated that
I used the January 2005 FAA registration database to determine the fleet size.
This is incorrect. In reality, my tool determines the *average* of the fleet
size from January 2002 to December 2004. It is this average that is used to
calculate the fleet accident rate.

There are other ways of making these determinations, but it by using the same
processes for both types of aircraft, the *relative* rates for each can be
obtained.

Ah, the heck with it...on to the data:

Cirrus Accidents during subject period: 20 (over a 3-year period)
Late-model 172 Accidents: 103(over a 3-year period)

Average Cirrus Fleet Size: 783 aircraft
Average Late-model 172 Fleet Size: 1993 aircraft

Average ANNUAL Fleet Accident Rates:
Cirrus: 0.85%
Late-Model 172: 1.72%

However, here's an interesting point: The C172 is used for instruction, while
the Cirrus is not. The NTSB lists none of the 20 Cirrus accidents as occurring
during instruction, while 68 of the late-model 172s are so listed.

If you eliminate the instruction accidents from the late-model 172 accidents,
the 172 rate drops to 0.59% *if* you assume the same fleet size. However, to be
an honest comparison, the fleet size would have to address only those aircraft
flown for personal pleasure or business...an almost impossible task. In any
case, the fleet rate would be higher.

From this data, I don't think the Cirrus rate stands out excessively.

Ron Wanttaja

Jose wrote:

Doing a little math:

And in the past year, the numbers have gotten worse.
Accidents and incidents (from theFAA and NTSB databases)
1998 2000 2001 2002 2003 2004 2005 2006
SR20 1 0 2 3 0 1 2 7
SR22 - - 2 2 3 8 12 15

TOTAL 1 0 4 5 3 9 14 22
rate (%): 50 0 2 1 .3 .6 .7 1
fleet size: 2 7 206 514 902 1491 1949 2323
SR22 fleet 121 383 687 1180 1560 1848
SR20 fleet 2 7 85 131 215 311 389 475

So, each year a bit less than one percent of the fleet bites it. The
rate seems to be increasing slightly in the last few years, but the
sketchiness of this data precludes a conclusion based on that.

It is good to see some fairly complete data. I agree that the
statistics are such that you can't draw a lot of conclusions as yet, and
when the fleet size was less than 500 it is especially troublesome as a
couple of crashes has a large affect on the percentages. However, as
the fleet has grown beyond 1000 and the rate is increasing nearly
linearly, that is something to be concerned about, in my opinion.


Matt

"Peter Duniho" wrote in message
...
"Kyle Boatright" wrote in message
. ..
Any aircraft has a baseline accident rate. I think the Cirrus has a
higher accident rate because a handful of pilots get themselves into a
mindset where they are willing to enter conditions they would have not
entered without the big round "insurance policy". Often they get away
with pushing things. Sometimes they don't, and those accidents are the
ones that are taking the Cirrus accident rate to higher than predicted
levels.

The problem is with the pilots, not the airplanes.

I've yet to see anyone document an accident rate that is actually higher
than might be expected (never mind "predicted"...who has predicted a
specific accident rate for the Cirrus, and why should we believe that
prediction?).

A quick NTSB database search shows in the last six months 4 accidents (2
fatal) involving a Cirrus SR20, and 52 (5 fatal) involving a Cessna 172.
The SR22 was involved in 7 accidents (2 fatal), while the Cessna 182 was
involved in 36 (6 fatal).

One might say that the fatal accident rate seems disproportionate (50% of
the SR20, 25% for the SR22 versus 10% for the 172 and 20% for the 182),
but at the sample sizes present, there's absolutely no reasonable way to
draw any valid statistical conclusion (and note that for the SR22 and the
182, the rates are actually similar).

Apples and oranges. The 182 fleet is many times larger than the SR22 fleet.
And the 172 fleet is near infinite compared to the Cirrus fleet. The
numbers look pretty bad for Cirrus.

"Dave Stadt" wrote in message
...

"Peter Duniho" wrote in message
...
"Kyle Boatright" wrote in message
. ..
Any aircraft has a baseline accident rate. I think the Cirrus has a
higher accident rate because a handful of pilots get themselves into a
mindset where they are willing to enter conditions they would have not
entered without the big round "insurance policy". Often they get away
with pushing things. Sometimes they don't, and those accidents are the
ones that are taking the Cirrus accident rate to higher than predicted
levels.

The problem is with the pilots, not the airplanes.

I've yet to see anyone document an accident rate that is actually higher
than might be expected (never mind "predicted"...who has predicted a
specific accident rate for the Cirrus, and why should we believe that
prediction?).

A quick NTSB database search shows in the last six months 4 accidents (2
fatal) involving a Cirrus SR20, and 52 (5 fatal) involving a Cessna 172.
The SR22 was involved in 7 accidents (2 fatal), while the Cessna 182 was
involved in 36 (6 fatal).

One might say that the fatal accident rate seems disproportionate (50% of
the SR20, 25% for the SR22 versus 10% for the 172 and 20% for the 182),
but at the sample sizes present, there's absolutely no reasonable way to
draw any valid statistical conclusion (and note that for the SR22 and the
182, the rates are actually similar).

Apples and oranges. The 182 fleet is many times larger than the SR22
fleet. And the 172 fleet is near infinite compared to the Cirrus fleet.
The numbers look pretty bad for Cirrus.

Did you adjust for the kind of flying done by each? No, you didn't.

moo

"Happy Dog" wrote in message
m...
"Dave Stadt" wrote in message
...

"Peter Duniho" wrote in message
...
"Kyle Boatright" wrote in message
. ..
Any aircraft has a baseline accident rate. I think the Cirrus has a
higher accident rate because a handful of pilots get themselves into a
mindset where they are willing to enter conditions they would have not
entered without the big round "insurance policy". Often they get away
with pushing things. Sometimes they don't, and those accidents are the
ones that are taking the Cirrus accident rate to higher than predicted
levels.

The problem is with the pilots, not the airplanes.

I've yet to see anyone document an accident rate that is actually higher
than might be expected (never mind "predicted"...who has predicted a
specific accident rate for the Cirrus, and why should we believe that
prediction?).

A quick NTSB database search shows in the last six months 4 accidents (2
fatal) involving a Cirrus SR20, and 52 (5 fatal) involving a Cessna 172.
The SR22 was involved in 7 accidents (2 fatal), while the Cessna 182 was
involved in 36 (6 fatal).

One might say that the fatal accident rate seems disproportionate (50%
of the SR20, 25% for the SR22 versus 10% for the 172 and 20% for the
182), but at the sample sizes present, there's absolutely no reasonable
way to draw any valid statistical conclusion (and note that for the SR22
and the 182, the rates are actually similar).

Apples and oranges. The 182 fleet is many times larger than the SR22
fleet. And the 172 fleet is near infinite compared to the Cirrus fleet.
The numbers look pretty bad for Cirrus.

Did you adjust for the kind of flying done by each? No, you didn't.

moo

The flights all involve an equal number of takeoffs and landings only some
are more successfull in the landing department than others.

"Dave Stadt"
One might say that the fatal accident rate seems disproportionate (50%
of the SR20, 25% for the SR22 versus 10% for the 172 and 20% for the
182), but at the sample sizes present, there's absolutely no reasonable
way to draw any valid statistical conclusion (and note that for the
SR22 and the 182, the rates are actually similar).

Apples and oranges. The 182 fleet is many times larger than the SR22
fleet. And the 172 fleet is near infinite compared to the Cirrus fleet.
The numbers look pretty bad for Cirrus.

Did you adjust for the kind of flying done by each? No, you didn't.

The flights all involve an equal number of takeoffs and landings only some
are more successfull in the landing department than others.

Unless you wish to redefine "flight" , no, they don't. Are circuits
"flights"?

moo