What's wrong with Cirrus Pilots?

[denverpilot]

Dick Collins is stirring the pot...

http://www.airfactsjournal.com/2012...s/?utm_source=facebook&utm_medium=socialmedia

 

[Jim Logajan]

His blog entry is so well planned out that he should be able to get a second one out of it by changing the title to "What's wrong with twin engine pilots?" and then making suitable substitutions. Second engine for parachute, and so on.

Perhaps his comment about doing things in a twin that one wouldn't do in a single is a bit of foreshadowing?

 

[Jim Logajan]

Another pilot and myself had a conversation via PM using an article with published data questioning the Cirrus safety record.

If you are referring to our message exchanges, the information involved was from a recent Aviation Consumer magazine article here (not available for free.) I've created the following table from some (but not all) of the entries they provided in their article:

[row][cell]Aircraft[/cell][cell]Overall Accidents per 100,000 hrs[/cell][cell]Aircraft[/cell][cell]Accidents with Fatalities per 100,000 hrs[/cell][/row]

[row][cell]Diamond DA40[/cell][cell]1.19[/cell][cell]Diamond DA40[/cell][cell]0.35[/cell][/row][row][cell]Cirrus SR22[/cell][cell]3.3[/cell][cell]Cessna 172[/cell][cell]0.45[/cell][/row][row][cell]Cessna 182[/cell][cell]3.3[/cell][cell]Cessna 182[/cell][cell]0.69[/cell][/row][row][cell]Columbia/Corvalis[/cell][cell]3.9[/cell][cell]Columbia/Corvalis[/cell][cell]1.0[/cell][/row][row][cell]Cessna 172[/cell][cell]4.3[/cell][cell]GA Average[/cell][cell]1.2[/cell][/row][row][cell]GA Average[/cell][cell]6.3[/cell][cell]Cirrus SR22[/cell][cell]1.5[/cell][/row][row][cell]Mooney M20[/cell][cell]6.6[/cell][cell]Mooney M20[/cell][cell]1.9[/cell][/row]

Of 31 CAPS deployments, they found that 12 of the aircraft had been repaired and returned to service.

Deployments as low as 300 ft and up to 187 knots have been successful. But one accident with deployment at 270 knots failed. Cirrus says that 920 feet is needed for full canopy opening and demonstrated deployment at 133 knots.

Aviation Consumer analyzed all the Cirrus accidents and estimates that nearly half the fatal accidents might have been prevented by CAPS deployments.

Oddly, no one seems to ask "What's wrong with Mooney Pilots?"

 

[Jim Logajan]

Where is the NTSB data for the above statistics. After all only the NTSB tracks all the aviation accidents and not Aviation Consumer. I found in the past Aviation Consumer to be in gross error on their assessments. Unlike the FAA or the NTSB Aviation Consumer does not have the resources or knowledge to make any proper evaluation of aviation products.

José

According to the Aviation Consumer (AC) article, the count of fatal and non-fatal accidents comes directly from over 500 NTSB records. Most people running such statistics get a count of accidents by using NTSB records. The problem is in the numbers used to normalize the accident counts. AC estimated the fleet hours for each airplane type by taking the average annual hours flown per airplane, as given by the Aircraft Bluebook Price Digest (ABPD), and multiplying by the number of registered aircraft of that type as determined by the GAMA and the FAA registry.

APBD in turn collects hours reported flown from information made available during actual sales reports.

Cirrus has made available its own estimate of hours flown for its airplanes by using "warranty claims, parachute repacks, and other methods" - so AC used the Cirrus numbers as a cross-check against their own method of estimating hours flown. They found the two ways of estimating yielded numbers within 4% of each other.

 

[Jim Logajan]

Besides, NTSB data is full of errors. There is no record of N273TE crash, for example. Airplane destroyed, 2 people died, but the only source of data is AirSafety wiki.

Unless the errors in the NTSB database favor one type of aircraft over another, such errors do not matter when one is simply doing comparative analysis (i.e. a simple ratio.) The errors are only significant when one is interested in getting at the absolute magnitude of the accident rate.

And as I point out in my last post, fleet hours are difficult to come by; the best that can be done is statistical sampling. And again, so long as the same methodology is used, one should still be able to use such estimates with reasonable confidence when doing comparisons.

 

[Jim Logajan]

Interesting point. Cirrus with a young fleet probably does fly many more hours on average vs. the piston GA norm. No one spends $800K to leave it in the hangar. Other types that have been made for 30-50+ years have a lot of un-airworthy tie down queens that probably skew their numbers toward better safety because they never leave the ground.

As I understand the description of their methodology, they seemed to limit the accident stats to only those aircraft built since the 1980s. So while that is a legitimate concern, it doesn't seem to be operative.

(At some point I should simply suggest users go pay for access to the article. It was written by this guy, who has enough epaulettes to be pretty darn authoritative