I know that some readers don’t like for us to write about general aviation accidents and safety. That is too bad because the best flying lessons are found in accident reports where we learn about the all-important things that we should not do in airplanes.
By my count there were 121 certified fixed-wing airplanes (excluding agricultural) in fatal accidents in the 48 contiguous states in 2013. There were 40 in experimental airplanes. Looking back at 2006, before the wheels came off most everything, there were more: 192 certified and 66 experimental. Does that mean flying has gotten that much safer? No. All it means is that GA pilots are flying that much less. The fatal accident rate per 100,000 hours has not changed much, if at all, in recent years.
For this discussion, I read all the 2013 accident reports. True, 2014 would have been more current but too many NTSB reports from that year are still preliminary. I think 2013 well represents what is going on now. It is also true that I could have done this with computer runs but I find I get a more accurate feel for what is actually going on by reading the reports. I have been doing this since 1958.
In doing safety studies, I have always concentrated on fixed-wing airplanes in the 48 contiguous states with agricultural flying not considered. That covers what most of us do.
Is anything changing? Yes, some things are and others are staying the same. Perhaps the biggest change is the dramatic drop in flying activity that has led to fewer accidents.
The fatal accident rate has been pretty stable in recent years at just over one per 100,000 flying hours. It moves a little but probably always stays within the margin of error on the estimated flight hours (by the FAA) which is little more than a wild guess.
Nobody will argue that this rate is acceptable. It is not, it is terrible, but it is what we get from our pilot population and the only way to change it would be to alter the behavior of pilots and that’s not going to happen. It would take Draconian legislation to improve the accident rate and that would effectively eliminate GA as we know it.
The definition of non-fatal accidents is so broad that many events the public and even pilots might think of as an accident don’t make it to the NTSB list because they don’t meet the parameters. Fatal accidents are absolutes and are also what we most want to avoid.
Over the years I have lost over 50 people I know, some well, in airplane accidents. When I study the subject, I like to think I am looking for something I might have said to them before they took off on their last flight. Forget about improving the accident rate because all that is really possible is to help individual pilots learn to avoid the things that establish that accident rate.
Starting at the beginning, before and during the first minutes of flight, the accident records make strong suggestions
First might be fitness to fly. That is an individual decision, made on the ground, and where the NTSB mentions pilot ailments and over-the-counter and prescription drugs in narrative discussions, word on this doesn’t always make it to the probable cause statement presumably because the NTSB didn’t think it was part of the probable cause.
Some of the stuff pilots take before flying is pretty scary and in reading the accident reports I often feel that the drug might have had a substantial impact on the pilot’s performance even if the NTSB didn’t think so.
I found 16 accidents in 2013 that were or could have been related to legal drugs. That covered about ten-percent of the total fatal accidents in fixed-wing airplanes so it is definitely a factor.
As in years past, impairment caused by either alcohol or illegal drugs is a factor in some accidents but there is no trend in either direction on this.
Once a pilot decides he is fit to fly, next comes the plan for flying. If anything, the trend here suggests that more pilots are taking off before they are fully prepared to do so. Perhaps that is because airplanes are becoming more complex.
A loss of control soon after takeoff shows up in ten-percent of the fatal accidents. Not much time is spent in the first few minutes of flight so that is a proportionately large number. These were split about evenly between VFR and IFR departures.
With high-tech the rage, I just naturally looked for something I thought might be there: a pilot’s lack of familiarity with the avionics system. Sure enough, the NTSB listed that in the probable cause of a 2013 King Air 200 accident. The pilot was experienced in the type but was flying an airplane with an avionics system that he had not used before. He stalled the airplane at 400 feet on the departure and the NTSB thought that a lack of familiarity the avionics had something to do with this.
Having all those whistles and bells in the cockpit is wonderful, but if you don’t know how to blow the whistles and ring the bells, distractions caused by that stuff can kill you quick.
There’s an interesting thing about the 2013 losses of control soon after takeoff in instrument meteorological conditions: five of the seven airplanes involved were turbine-powered. Four of these were singles. Is this an aberration or a new trend?
Accidents like this used to be the province of single-engine retractables and light twins. These airplanes are far less active now and some of the pilots who used to fly them are flying turbines today, if they can afford to do so. So I think this a matter of the same type pilots flying different airplanes.
Question to self: “Am I sure I have all the ducks in a row for this departure?”
When considering certified airplanes only, engine failures and power problems loom large as an increasing area of risk. Close to 20-percent of the 2013 fatal accidents fall in this category. Almost half of those were multiengine airplanes. It has always been true that your chances of being done in are greater in a multiengine airplane after an engine failure than they are after the engine fails in a single. The flat-earth folks will give you many reasons why that shouldn’t really count but the facts are the facts.
What has changed is that half the twins with power problems were turbines. That is something new and might be attributed to the turbines flying relatively more hours as time passes. It is true that turbines have far less catastrophic failures than pistons but a lot of other things, including incorrect pilot actions, can interfere with the power output of turbine engines.
As always, stall/spin accidents are number one on the list, accounting for about 25-percent of the total fatal accidents. I’ll talk more about weather accidents in a minute, but here I want to say that they do have something in common with stall/spin accidents: both are usually fatal. Sure, people have survived both accident types but they are in a minority. In fact, a higher percentage of the occupants in midair collisions survive than do stall/spin and weather accident participants.
There is something new in the stall/spin reporting in 2013. Where the NTSB used to often say that the pilot failed to maintain airspeed, angle of attack is creeping into the discussion as they mention the failure to manage angle of attack as a cause.
It’s almost as if someone had an “aha” moment about angle of attack. News flash: knowledge of its importance has been around since the Wrights’ first flying, instrumentation has been available for years (I had AOA instrumentation in my Pacer 60 years ago), and in fact, most airplanes flying have an angle of attack warning system. It has always been called a stall warning but what it tells you is that the angle of attack is nearing that which results in a stall. It warns you of that regardless of the indicated airspeed, as in an accelerated stall.
Maybe if the stall warning were renamed the AOA warning perhaps the near-hysteria among government folks and some safety mavens about AOA would go away.
When my father started AIR FACTS in 1938, stall/spin accidents were the safety subject of the day. They still are and that will likely remain true for a long time. The accidents of today bear a great similarity to the ones of 77 years ago and I honestly can’t read the accident reports and identify many, if any, accidents that more complete AOA instrumentation would have prevented. If a pilot can’t get the message from the airspeed, from feel of the airplane, from the look of what is going on, and from the bleat of a stall horn, how can another gauge on the instrument panel help?
If there is a standard stall/spin event in more or less normal operations, it comes out of a steep turn at low altitude. Unless a pilot is cutting up, the most likely place for that to occur is on the turn from base to final in a badly botched approach. When you look at such accidents, the angle of bank in those fatal turns is usually far in excess of 30-degrees, which is the guideline for maximum bank angle at lower altitudes. You can bet that a pilot who is really bending it around in a steep turn to final is not looking at the instrument panel.
If the weather accidents that occur in visual and instrument conditions are combined, then in 2013 they actually outnumbered stall/spin. That would make weather the number one bad thing but weather accidents have always been separated into those that occur on VFR flights and those that occur on IFR flights. As has been true for a while, they are about evenly divided though the trend has been toward more IFR and fewer VFR weather accidents.
Years ago, when most weather accidents were VFR, many of us thought that promoting instrument flying could result in a better safety record. We were wrong. It just moved the accidents from one column to another. Instrument operations are far more difficult than visual and average GA pilots have not been able to master this as well as we thought they could.
The number of IFR accidents at night is completely out of proportion to the amount of flying done under those conditions. It is generally agreed that maybe ten-percent of the IFR hours are flown at night and more than 30 percent of the IFR accidents happen at night. If anything, that is getting worse with the passing of time.
Night IFR is just not something that the non-professional pilot does often so there is no way for him to be really proficient at it. Is the huge increase in risk really worth it? If there is a real need to fly night IFR then there is also a need to do IFR training and proficiency flying at night.
Airframe failures in flight are truly bad news and they have been and remain a factor every year. They occur as a result of a high-speed loss of control as might happen when a VFR pilot flies into clouds. It can also happen when an instrument pilot becomes distracted and loses control or when a pilot out pranking around lets the airplane get away from him. The result is flight outside the envelope for which the airplane was designed, with both the speed and g-load limits exceeded.
This type accident used to happen mostly in V-tail Bonanzas, Cessna 210s and Piper PA-32 retractables such as the Saratoga. No more. I counted 11 airframe failures and the only certified airplane to show up in more than one was the Piper PA-34 Seneca twin with two.
Twins used to be a rarity in airframe failure accidents so this is unusual. I would say it was a quirk. There was one V-tail on the list, one Cessna 210 and one Piper PA-46 Malibu. Four of the 11 were experimentals. That’s ten-percent of the total experimental airplane fatal accidents which is higher than for certified airplanes.
Ice, thunderstorms and midair collisions have decreased in number by quite a bit more than flying activity. I would imagine that is because pilots have learned to blow the whistles and ring the bells of electronic weather and traffic information systems. At least I hope that is true.
One other flying technique area that comes up short is the missed approach or go-around. The inability of pilots to handle this accounts for ten-percent of the fatal accidents. This clearly illustrates a need in the training process.
Nobody expects that they will miss an approach or have to fly a go-around even though a pilot should always be locked and loaded for either event. Couple that with the fact that anything that is done so seldom tends to take a back seat to other things. That needs to be addressed in training and practice.
When the list of certified airplanes involved in fatal accidents in 2013 is examined, a few things stand out. There are big changes but, with one exception, the changes are likely related to less flying in particular types.
The exception is the Cirrus. Where there used to be a bunch of Cirrus airplanes on the list every year, I found only two for 2013. I think this does show that helping individual pilots avoid the bad stuff is the way to go. Certainly there has been great effort put into helping Cirrus pilots operate their airplanes more safely and I think this is cirrus accidents.
On the other hand, there needs to be an asterisk by this. While there were only two Cirrus fatal accidents in the U. S., there were a total of eight worldwide. That does put a different light on the subject but I don’t exactly know what it means. Maybe the Cirrus group knows of a reason why the airplane is operated more safely in the U. S. than elsewhere.
The pressurized piston singles, the Cessna P210 and Piper PA-46 used to appear frequently on the accident roster. In fact, at one point I calculated that the P210 had the worst accident rate in the fleet with the PA-46 almost as bad.
No more. There was one PA-46 on the 2013 list and no P210s. In the case of the P210, I suspect that is because the airplanes don’t fly nearly as much as they used to. I retired my P210 eight years ago and, even back then, properly maintaining the airplane and keeping it ready for reliable service was a frustrating and almost frightfully expensive proposition.
The PA-46, which is still in production where the P210 is not, probably does better because of an increased emphasis on training and perhaps because pilots have come to understand that pressurized singles can be dangerous as all get-out if not used carefully and correctly.
One jet, the Premier (Raytheon 390), was in three fatal accidents in 2013. Two of the three were apparently flown by owner-pilots.
I mentioned not using computer runs. The Premier gives a good illustration of why this can lead to errors. Query the NTSB database for “Raytheon 390,” which is the official designation, and only one comes up. The other two are listed as Hawker Beechcraft 390 and Beech 390. However a “model” query of just “390” will bring up all three if anybody who is looking knows that a Premier is a 390.
Among older airplanes, there were four Piper Twin Comanches which would be high for the likely number of hours flown by that fleet.
So, the general aviation safety picture continues to evolve as the fleet and the amount of flying continues to change. And anybody who suggests that some technological wonder will bring much change to GA safety flies with rose-colored glasses. It is quite obvious that keeping the head containing the brain in the left front seat well separated from the butt in that seat counts for more than anything on or in the instrument panel.