Accidents by type review:

[Let'sgoflying!]

@wanttaja
I think this site (below) provides a decent world-wide review of accidents by type.
I want to review the Daher Kodiak100(1) & compare it to the average, and probably some other specific types.
Overall, I think it looks like a pretty safe aircraft.
If you only use it on land, remove all the water accidents it looks even better.
What do you think?
I think between the companies that built them, there were around 1500 made in the last 17 years (some sites say 20).
This site shows 34 accidents with 10 fatals.

Aviation Safety Network > > ASN Aviation Safety Database results

Aviation Safety Network: Aviation Safety Network: Databases containing descriptions of over 11000 airliner write-offs, hijackings and military aircraft accidents.

asn.flightsafety.org

 

[wanttaja]

Not really enough accidents to come to any conclusions. I usually require 50 US accidents in the NTSB database; there's only 12, and several of them are foreign accidents for which detailed information and probable cause is not recorded. The ASN list includes incidents, which are not included in the NTSB accident database (Example: N307RW). My database also only goes through 2022.

I've attached a PDF with the Kodiak accidents from the NTSB database, 2008-2022.

Ron Wanttaja

 

[dans2992]

Keep in mind, the Kodiak is likely used in some “higher risk” types of flying (ie unimproved strips, off-airport, back country, etc.)

 

[wanttaja]

Keep in mind, the Kodiak is likely used in some “higher risk” types of flying (ie unimproved strips, off-airport, back country, etc.)

Certainly; of the 11 US Kodiak accidents that listed a landing gear type, five were either floats or amphibious floats. Accidents didn't necessarily happen in the water, but it's an indication of the use of the aircraft.

Keep in mind, too, that there are also more-experienced pilots involved.

For instance, I get a median pilot experience of 250 hours for Cessna 172 accidents, and 1,000 hours for homebuilt accidents. The median for US Kodiak accidents is 4900 hours (albeit only a 10-accident set). Two of the ten accidents were flown by Private pilots, the rest all had Commercial or ATP ratings.

Median time-in-type was 99 hours.

Ron Wanttaja

 

[Rockymountain]

5 fatal accidents out of 300 built, for a turbine? Doesn’t sound overly safe to me. Looking at those 5 fatal accidents, all occurred near an airport on take off or approach phase, only one on floats. These don’t sound like high risk flights to me. Especially if professional operations, with professional pilots, where a safety management system should be in effect.

 

[StraightnLevel]

Keep in mind, too, that there are also more-experienced pilots involved.

For instance, I get a median pilot experience of 250 hours for Cessna 172 accidents, and 1,000 hours for homebuilt accidents. The median for US Kodiak accidents is 4900 hours (albeit only a 10-accident set). Two of the ten accidents were flown by Private pilots, the rest all had Commercial or ATP ratings.

Median time-in-type was 99 hours.

Slightly off-topic, but this brings a question to mind:

I've seen a lot of data quoted talking about the 200-500 hour PPL "danger zone" where pilots are more likely to be involved in an accident. I also see a reasonable number of accident reports that involve very high-time experienced pilots.

So the question: Has this data ever been adjusted for the number of pilots at a given experience level? At any given time, how many pilots - what share of the group - are in the 200-500 hour segment? How many below 100? How many between 500-1000, 1000-2000, 2000-5000, etc.? It strikes me that there this is inherently a bottom-loaded group, as there will always be more people in the lower hour groups, and many will never reach the higher levels.

The reason I ask is statistical. It's one thing to note that there are more accidents involving a certain type of plane, certain pilot rating/experience level, etc. This has little value, though, taken out of context of the sample set size. When looking at planes, we know that the number of accidents must be divided by the number of existing airworthy examples, but I haven't seen the same sort of breakdown on pilots. Is it out there, and I just don't know where to look? Is it even possible to collect outside of insurance data, given the way in which hours are logged?

 

[wanttaja]

5 fatal accidents out of 300 built, for a turbine? Doesn’t sound overly safe to me. Looking at those 5 fatal accidents, all occurred near an airport on take off or approach phase, only one on floats. These don’t sound like high risk flights to me. Especially if professional operations, with professional pilots, where a safety management system should be in effect.

From my data, there were 12 US accidents (through 2022), with two incurring fatalities. That's a ~16% fatality rate; pretty typical. It's less than the Cessna 210, it's more than the Cessna 172.

NTSB found that both the fatal Kodiak accidents were due to spatial disorientation. Both pilots (one Commercial, one Private) had 3400+ total time.

Ron Wanttaja

 

[wanttaja]

Slightly off-topic, but this brings a question to mind:

I've seen a lot of data quoted talking about the 200-500 hour PPL "danger zone" where pilots are more likely to be involved in an accident. I also see a reasonable number of accident reports that involve very high-time experienced pilots.

So the question: Has this data ever been adjusted for the number of pilots at a given experience level? At any given time, how many pilots - what share of the group - are in the 200-500 hour segment? How many below 100? How many between 500-1000, 1000-2000, 2000-5000, etc.? It strikes me that there this is inherently a bottom-loaded group, as there will always be more people in the lower hour groups, and many will never reach the higher levels.

The reason I ask is statistical. It's one thing to note that there are more accidents involving a certain type of plane, certain pilot rating/experience level, etc. This has little value, though, taken out of context of the sample set size. When looking at planes, we know that the number of accidents must be divided by the number of existing airworthy examples, but I haven't seen the same sort of breakdown on pilots. Is it out there, and I just don't know where to look? Is it even possible to collect outside of insurance data, given the way in which hours are logged?

The problem is, we don't know how many pilots at a given experience level are still active. Maybe the accident rate decreases for pilots with more than 20,000 hours, but if most of those people are not flying any more, it skews the results.

I did attempt to look at the accident causes for homebuilt accidents involving pilots with 100 hours, 1,000 hours, and 20,000+ hours.

Homebuilt Accidents: Captains vs. Fledglings - KITPLANES

How does pilot experience affect the accident rate?

www.kitplanes.com

Ron Wanttaja