Cirrus Fatalities Have Critics Questioning Safety

Your use of statistics is pretty reckless (which is ironic for an PI lawyer).

Before quoting statistics such as "Using those parameters, COPA claims Cirrus aircraft have a fatal accident rate somewhere between 1.42 and 1.76 per 100,000 hours flown...compared to the overall general aviation rate of 1.19 fatal accidents per 100,000 hours flown" you should really consider learning how to test the significance of such numbers. If a single Cirrus accident occurs how much does it change the Cirrus rate? What if a normal general aviation plane crashes? How much does that change the general aviation (GA) rate?

Also, do you even know what types of planes are covered by the term GA? Does it include private jets with two pilots holding Airline Transport Pilot ratings? Do you even know what that means?

You might also investigate the rumor that the Cirrus' will spin easily, and that spin recovery is based on deploying the parachute, rather than standard pilot flight control recovery procedures. The parachute is ineffective at low altitudes. FAA apparently approved this parachute based spin recovery procedure during the Cirrus' initial airworthiness certification.

My pilot friend was killed last year, riding as a passenger in a Cirrus in Oklahoma, that spun at low altitude during landing approach. The Cirrus is not as safe as it could be.

You might also investigate the rumor that the Cirrus' will spin easily, and that spin recovery is based on deploying the parachute, rather than standard pilot flight control recovery procedures. The parachute is ineffective at low altitudes. FAA apparently approved this parachute based spin recovery procedure during the Cirrus' initial airworthiness certification.

My pilot friend was killed last year, riding as a passenger in a Cirrus in Oklahoma, that spun at low altitude during landing approach. The Cirrus is not as safe as it could be.

I flew over 500 hours of Part 135 PIC time in a Cirrus. There are enough pilots like myself who agree that the only time one should use the parachute is at night, in IFR conditions, full electrical malfunction, engine loss, and over the mountains. When would this occur? Never, which means the parachute should only be used as an absolute LAST resort. Truth be known, most if not all operators of this machine don't have experience in it when they fly it for the first time (no matter how many hours they have total). And training doesn't reflect real world. The best way to learn without killing oneself is to simulate. Why doesn't the industry stress more simulator training for low experience/low time pilots. These accidents are occurring because of "guessing what the aircraft is hyped up to do" versus "what it actually does". Go back to square one....Training, as opposed to the underground philosophy of "Parachute as a first resort"

Brian,
Thanks for commenting, but I wish you would have read the article more carefully. I am hardly being reckless with statistics. Fatal accidents per 100,000 hours flown is the method used by the FAA, other federal agencies, the AOPA, and every other alphabet group I know in computing fatal accident rates.
The statistics cited in the article came from the NTSB database and were compiled by the Cirrus Owners and Pilots Association, which was duly noted.
If you go back and re-read the article, you will see that the GA accident rate for the time period studied was 1.19 fatal accidents per 100,000 hours flown including all GA aircraft (and yes, I do know what types of aircraft are covered) but that rate climbed rather substantially, comparatively, to 1.86 (which was higher than Cirrus rate) per 100,000 hours when multi-engine aircraft and private jets were factored out, and only single engine piston aircraft stats were used (and yes, that would be factoring out those prviate jets with the ATPs.)
Thanks for your input.

How does one recover this aircraft if it didn't have a parachute?? I am going out on a limb--don't have access at this moment to any pilot handbook for SR-22, how did it get certified for spin testing? Pilot deploys parachute, stops spin, then releases the parachute? nah impossible... or rides parachute down to impact?

Well, answers to my previous question -- "Once in a spin the SR20 and SR22 are virtually impossible to recover, according to the test pilots. Remember that spin testing in certification is done with a special tail parachute for breaking the spin that can then be cut away inflight." Yup, this sounds quite unhealthy to normal flying pilots...I stand corrected...

Where did the quote "Once in a spin the SR20 and SR22 are virtually impossible to recover, according to the test pilots." come from?

While the parachute was used in lieu of spin testing in the U.S., the SR20 and SR22 did pass spin testing for European certification. When I trained at the Cirrus factory, my instructor said that test pilots said that normal spin recovery techniques work with the Cirrus, but that it "needed to be done right" which left me with the impression that there was less room for imperfect technique.

I think the biggest issue is low time in type. I think most accident victims had less than 100 hours in type, which is common for all GA accidents. Interestingly, a disproportional number of accident victims were not members of COPA, which suggests that they were a little less interested in safety to begin with, or perhaps lacked the curiousity to find out that there is an organization that helps members become familiar with their airplanes' characteristics.

I have owned 2 Cirrus aircraft and have logged several hundred hours in them. There is much more that can be said in this debate, but I want to make (remake) the point that every accident of Cirrus aircraft can be studied individually and in some detail by anyone interested. The truth is that such a study reveals that most of the accidents were (as is the norm in aviation) pilot error. There is not much speculation needed to understand the real Cirrus safety record, though you wouldn't know it from some of the posts here. The Cirrus flies with very crisp and predictable response. It is a high performance aircraft and demands more skill than a 172 or 182 to be safe if for no other reason than things happen more quickly at higher speeds. But it is by no stretch difficult to fly. One predictable truth with ANY aircraft is that low speed, close to the ground, steep banking WILL KILL YOU. No amount of spin training or aircraft capability will save you from a stall/spin turning base to final.

I have about 700 hours on my Cirrus and believe that for equivalent cross country missions, this aircraft is safer than the average general aviation airplane. That being said, anyone who has been in the Cirrus community for any period of time, will tell you that the Cirrus is on average flown on more demanding missions than most other general aviation planes and can be expected to be involved in more "risky" flights. Let's be realistic, general aviation is dangerous, as is, scuba diving, horseback riding, skiing, climbing, motorcycle riding, bicycling and so many other fun activities.

an owner's review by Philip Greenspun, ATP, CFII, in July 2005, updated April 2009----"Once in a spin the SR20 and SR22 are virtually impossible to recover, according to the test pilots." The author was mentioning the sleekness of this aircraft and that this aircraft shouldn't be flown slow. I'd love to get my hands on a Cirrus--only after I had training and time in though.

One person brought up spins. First, most stall/spin deaths are in the pattern at low altitude where recovery from a spin is next to impossible in any aircraft. A base-final turn is usually 500' above ground. Secondly, twin engine aircraft are not spin certified. That 737 you get in isn't spin certified nor is any other twin that I know of. Cirrus aircraft use a NASA wing design as do the Columbias that is meant to be easily recovered from a stall and thus help avoid entering a spin. Accidents do NOT show spins as a problem. There was at least one base to final stall but data recoveed from the aircraft showed that the pilot failed to keep airspeed adequately high i.e. he let the plane get very slow.

Small aircraft accident statistics are skewed heavily by a few accidents. For example, in 2007 the Cirrus SR22 had a better record than the Cessna 182. Also, different planes have different missions. The Cirrus aircraft should be compared to Mooney and Bonanzas. If you average 2007 and 2008 then Cirrus is better than the Bonanza and worse than the 182 and Mooney. The fact is that the Cirrus is fast and comfortable so it encourages use for long trips and as a business tool.

Kudos to Lisa Brown:
As a 2300 hour instrument pilot who owns a T210 and has a direct family member with an SR22-Cirrus,
I have read numerous NTSB accident reports on Cirrus and think that pilots take unnecessary chances in them because of the parachute. It would be nice if Lawyers could be prevented from flying them, because the end result will be increased insurance rates for all Cirrus owners.
Once again, very good article Lisa Brown.

So one begs to ask...is the problem too many bloody rich pilots that have more money than common sense? Or is the aircraft the problem?

I suspect that those with which are well-to-do are used to buying the best and want the best right away would never see the "loonacy" in actually buying a less capable aircraft (or being SEEN in something like a Cessna) to hone their skills and build hours BEFORE jumping into the Porsche of SEL GA aircraft. It's like giving a 16 yr old a Porsche for his first car. You KNOW they are going to wreck it.

Maybe the answer to Cirrus' problem is to not release ownership of the aircraft until the pilot has gone through a more thorough training process for the safety of themselves and everyone else?

Interestingly enough, you don't see this problem with aircraft like a Diamond DA40...which would seem to be a nice choice to start in before moving up to the higher performance models.

Personal aircraft ownership is all about an almost forgotten American concept called personal freedom.
Insurance companies have requirements and the government has requirements and the pilot has a host of choices to make before and during each flight.
Cirrus is less docile than a Cessna 182, but it is not really a difficult plane to fly either.
At the end of the day pilots have to decide what is safe. Taking off with 700 foot ceilings is a very risky proposition in any single engine airplane. A lot of intrument pilots would not do it.

I apologize to Max Trescott if this sounds like an ad hominem attack. Max is a famous aviation author and is as concerned with safety as any of us in the industry. That said, it is troubling to me that he apparently he bases his spin recoverable remarks on statements from his instructor rather than the Cirrus Pilot Operationg Handbook (POH) he presumably read as part of his training. From the 2009 Cirrus Turbo Perspective POH: In the emergency procedures, section 3 of the Cirrus POH, they state, Quote “the SR22 has not been tested or certified for spin recovery characteristics.” End Quote. ;They further state that the stall characteristics of the SR22 are such that it is possible for it to spin and that the pilot may not realize that the aircraft it has entered a spin; that the only demonstrated method of spin recovery is to deploy the parachute; that the parachute may not work at speeds in excess of 133 kts; that if deployed, the aircraft may descend at a rate of 1700 fpm; and finally, that the plane and occupants may be dragged along the ground by the parachute chute after impact.

As Max undoubtedly knows, the FAA spin recovery matrix is based on more than 500 points in an aircraft flight envelope (weight, cg, power setting, altitude, etc)that the manufacturer must demonstrate that the aircraft will recover from a fully developed spin in less than one additional revolution, using normal control inputs. One would more legitimately conclude from the manufacturer's statements in its owners manual there are points in the aircraft flight envelope in which it will spin and not recover.

I completely agree it appears the majority of Cirrus accidents are pilot error. We may speculate endlessly why Cirrus pilots appear to be more error prone, or whether it may be easier to lose control of this plane and harder to recover from loss of control. The fact remains, to the best of my knowledge, Cirrus is the only Part 23 OEM single engine piston aircraft manufacturer in the market today whose aircraft is not certified by the FAA to be either spin resistant or spin recoverable. It is instead, the only one required by the FAA to install a ballistic recovery chute in lieu of demonstrating that it is either spin resistant or spin recoverable, in order to be allowed to sell the aircraft.

Cirrus is an extraordinary aircraft marketer and has introduced to aviation persons who perhaps would never had considered learning to fly. Those of us in the industry are grateful and undoubtedly these persons are largely thrilled with all the fun and opportunity this has afforded.

I also think that accident statistics are finally dispelling the illusion that an airframe parachute does not create the panacea of safety that it has been represented to provide. Initial training, recurrent training, aeronautical decision making and risk management are paramount for us all.

Fly safe.

Just a couple clarifications: The FAA certifies a plane to be legal to do spins, or not legal to do spins. There is a difference between being certified and being tested. There are a number of four place GA aircraft in which spins are prohibited, which means they are not certified for spins. The Tiger AG-5B, the Mooney, and the Bonanza are just a few examples. Cirrus, on its website, says the aircraft wing design in which the inner part of the wing stalls first, rather than the tip, makes it spin resistant. The FAA did not "require" Cirrus to install a BRS in lieu of the demonstration of spin recovery. Rather, the FAA "accepted" the BRS as an equivalent level of safety (ELOS)

Let's line up some indisputable facts:

1. The Cirrus has been spin tested by the JAA/EASA. Nothing noteworthy was discovered.

2. Spins are not a significant factor in Cirrus accidents to date.

3. As documented in the article, Cirri have a better fatal accident rate (1.42-1.76) than the single piston engine aircraft class (1.86).

4. As even a cursory review of COPA discussions will show, the Cirrus flyer community is extremely concerned that its many safety features have not translated into a bottom line improvement.

Now for opinions.

1. We pilots are congenital soapboxers, so strong views for and against any particular aircraft are a dime a dozen. But it's troubling when myths ("it's spin prone") persist despite contrary evidence in our supposedly analytic demographic.

2. To this reader, the article is biased. Verbs used in reference to COPA: "argues", "admits", "concedes" (twice). Verbs used in reference to non-Cirrus parties? Glad you asked! They are: "reports", "say"/"says", "cited", and "consider". And then there are lazy constructions like "In talking with pilots, many will say that it sure seems like there have been an inordinate number..." Gee whiz! In talking with average folks many will say that Armstrong never landed on the moon, that the UN is flying black helicopters over America, and that psychics should help police find dead bodies. Why not buttress your future articles with such strong supports?

Having flown ( counting only 10+ hours) a 152, multiple 172s, a 182, a Citabria, an Archer, and for the last 500 hours, a Cirrus SR-20, I can say from experience that the Cirrus is easily the safest. The situational awareness its advanced avionics enable, its simplified low-stress operation, and the "last chance" chute, means it's the safest I've ever felt in a plane.

Flying, like many worthwhile activities, comes with risks. It's incumbent upon pilots to look to our tendencies and practices in order to attack the most likely point of failure. Only by ruthlessly questioning our own decisions, often shaped by expedience and self-delusion, can we be safer.

Journalism is not dissimilar in one sense. The possession of a megaphone - even a special-interest blog with narrow readership - should also force a writer to examine and overcome his/her biases to deliver a better reader experience.

I am unfamiliar with Ms. Brown's vouched-for oeuvre, but the tenor of this article suggests that she could benefit from at least as much self-examination as the typical Cirrus pilot.

I did an interesting exercize. I looked at Flightaware to get an idea of planes flying in the system. This is a good indication of how much a model is used for long trips and "weather" flying. I also looked at the NTSB database and counted fatal accidents for that model from 1/1/07 to 1/1/09. As a figure of merit I took accidents divided by aircraft in the system. I need to look at this over a period of several days but I wanted to see if anything stood out. Since one person on this blog flies a 210 I looked at that and several other serious travel planes (not trainers). Diamond is interesting both because of the excellent safety recorrd for the DA40 and the low use for IFR flights. Here are some numbers for types with a lot of planes on IFR or flight following:

Cessna 182 0.72
Cessna 210 1.46
Beech Baron 58 0.51
Mooney (all)0.55
Beech BE36 0.47
Beech BE35 0.71
Cirrus SR22 0.53

Lower numbers are better. If I can figure a way to automate this and smaple each hour for a week then it would be more meaningful but it is still useful. It measures serious use of aircraft as opposed to pattern work. It also shows actual flight usage rather than number registered. BTW Diamond had only one DA40 fatal in the NTSB database for the period used but also only had 4 planes in the system when I looked. That would give a great metric of 0.25 but the confidence interval is huge due to the small numbers. The 210 stands out as particularly bad so that will be interesting to track and see if it holds up over time. During the period in question the SR22 had 14 fatals listed while the 210 had 19. However, Flightaware showed over twice the numbers of SR22's in the air as 210's.

The 210 numbers I posted may give an incorrect impression. I hadn't initially done the 210 and thought the Flightaware page was static. I think it may have updaetd and, it now being late at night, the number would be much lower for C210's in the air. I'll need to make sure I do a fair comparison. More work to do. If you want to check what I am doing you can go to flightaware.com and click live tracking and then in the upper right ALL. For the accident data I use ntsb.gov and search the aircraft database for fatals (shown in red). While soem accidnets aren't in the system this count seems fair to all types. Use Cirrus (eliminate VK 30 and separate SR20 & SR22) and 210 and 182. You have to be careful to just use 182 to get all 182 varients.

To: Jane Akre & Lisa Brown:
Thanks for writing this interesting article that follows up on my initial article questioning the safety of the Cirrus vs. Cessna. The analysis from pilots has been interesting, ranging from blaming "type a" lawyers and doctors for crashing the Cirrus, to laying it on "pilot error." Hey, if doctors and lawyers weren't buyers of the Cirrus, the company's sales would slump!
I have read the posts with interest and my original premise about the cessna rates of fatalities head to head in flight time have not been refuted (the stats were raised by Steve Wilson's first blog, not mine). I think a recurring comment is the reliance on the parachute. Perhaps Cirrus needs to virtually mandate simulator training for new AND used purchasers. Cirrus could offer some rebate or warranty inducement or deal if the pilot will undergo a minimum simulator training for example.

Rick,

You state that no one has challenged the statistics. I do. The closest plane to the SR22 in terms of a Cessna would be the 210. The 182 is slower and with a lower stall speed. Whether you take the 210 or the 182 I don't see either being safer if you look at actual IFR flight data. As for Steve Wilson's blog, he is a Cessna salesman. Steve's data throws in the 172 which is a different class of airplane used as a trainer. Training flights are statistically much safer.