Could a good proxy for aircraft utilization be to crawl aircraft listings and extract total time?
Summing up the total time for each model listed for sale would give the total flying hours of that “fleet”.
This is a great idea, love the creativity!
Antonio one question in this aspect of the timespan of the accidents is: are the pilot skills different today than 70 years ago? Another aspect regards instrumentation: with glass cockpit, GPS and things like AOA indicators there’s much more aid to obtain from the panel than like 50 or even 70 years ago.
So if the accident rate lowered over the years due to better training or equipment, then the accident rates of old planes are not comparable over the whole time.
Antonio wrote:
ie is the 3328 TT average for Mooneys, which have been in production for 71 years, directly comparable to the 1601 hrs TT of DA40’s which have been in production for only 27 years?
Mooneys are mostly privately owned, rarely club airplanes, hence the relatively low TT most of them have.
@xavierde this should compensate average differences in utilization amongst aircraft types.
Would it still have to be corrected for length of the production run? ie is the 3328 TT average for Mooneys, which have been in production for 71 years, directly comparable to the 1601 hrs TT of DA40’s which have been in production for only 27 years?
Could a good proxy for aircraft utilization be to crawl aircraft listings and extract total time?
Summing up the total time for each model listed for sale would give the total flying hours of that “fleet”.
Our Jodel Syndicate started in 1984. There were 2 accidents before I joined in January 1990.
There were 3 accidents by 1999.
There were no more accidents until we sold the DR1050 in late 2020.
Result of increased experience of type in the Syndicate?
Mooney_Driver wrote:
IE the TB20 has a fatality rate of 69% (if I read this right, 69% of the reported accidents are fatal)
No, sorry I did not make that more clear: this only means that in 100 TB20 accidents there are 69 people dead (ie 69 fatalities).
Hence the suggestion to correct or divide by number of seats since a six seater is bound to fly with a higher number of POB than a four seater. If you do not do that, a 747 would be found to have a much much higher fatality rate than a TB20, (1666 fatalities for 158 accidents or 1054% fatality rate for 747-200 in ASN) which would be misleading. Dividing by 300/4 seats will result in 14% for 747-200 vs 69% for the TB20 which makes much more sense.
I have to admit I wonder about those numbers.
Some of the airframes which have horrible numbers have not been known as “death traps” or “Doctor Killers” throughout their service, yet they have horriffic fatality rates while accidents by serial number are very low. IE the TB20 has a fatality rate of 69% (if I read this right, 69% of the reported accidents are fatal) and the M20x had 74% of reported accidents include fatalities? At the same time, Accidents per SN are quite low for both of those, while accidents per SN (would that mean that x% of the fleet was involved in accidents?) is comparatively low.
The other question is how reliable the data of ASN are and how many accidents are reported vs actual numbers. I would suggest that particularly the fataliy rate might be massively increased over the actual accidents having taken place if accidents are not included in the ASN DB.
