#11: Some accident analysis for 2023

I’m constantly updating my spreadsheet on Part 135-involved accidents, trying to keep the clearest picture on the industry’s accident stats. Here’s what I’ve got for 2023:

(I just really like this photo of “women working in the 1950s”. Consider this a peek inside my head when I’m in the NTSB database. Also - see the “housekeeping” note at the end.)

There were 61 accidents in the U.S. that year for Part 135 operators, resulting in 32 fatalities and 13 serious injuries. Not all the final reports are out yet, but we can glean some trends from what is there. First, there was another accident for Guardian Flight. The fatality crash in Nevada in February followed fatality crashes in Hawaii in 2022 and Alaska in 2019. Altogether, eleven people were killed in those accidents and it is fair to question if there are systemic problems with the company. Only the Alaska crash has a final report but unfortunately, because it went into the ocean and not all of the wreckage was recovered nor was the pilot found, it is undetermined. We have to wait on the other two for a more complete picture of Guardian’s operation.

There were eight total medevac accidents in 2023 following eleven in 2022. Air Methods was the most frequent operator involved - it crashed six times in the two years. (In one of those cases the pilot tested positive for cocaine.)

Medevacs can be tricky to track because if the aircraft is empty before it picks up a patient or after transport, an accident occurring in those legs can be classified as Part 91 under “positioning”. (This was the case for the Alaska Guardian Flight crash.) When I write specifically about air ambulance ops, I’ll get more into how this distinction is made but suffice to say, to capture all the medevac accidents you have to look both in Part 135 and Part 91 listings in the NTSB database.

A note on my research methods

I look through Part 91 accidents every month to be sure and capture Part 135 accidents, like medevacs, that might have occurred during training, repositioning, business, etc. That’s why my analysis is framed as “Part 135-involved” and not just “Part 135”.

Now back to 2023 accidents

Moving on…it’s proved impossible to see any kind of industry-wide correlation between flight time and accidents in part 135. For 2023, the flight time range is huge - from 770 hours for a Yute Commuter Service pilot in AK to 19,269 for a First Class Air Charter pilot in FL. I don’t have flight times for all the pilots yet (still waiting on those final reports), but it tracks that time is not a reliable predictor for an accident. I have 30 years of accidents in my Alaska database and the flight time is all over the place, so while I collect it and it does matter (sometimes a lot) for some accidents, overall it’s not a primary causal factor.

It will likely come as no surprise that 20 of the accidents last year occurred in Alaska, accounting for eleven fatalities. While the typical explanation for this is that Alaskans are more dependent on aviation and the conditions are “more dangerous” which equals more accidents, I will point out yet again that the state’s navigational infrastructure is woefully underfunded and undeveloped compared to the Lower 48 and FAA staffing is subpar to say the least. (Hopefully the funds in the FAA Reauthorization Bill will finally affect positive change for certified weather and IFR accessibility.) (I really don’t care that this bill says it’s going to “reduce the rate of fatal aircraft accidents by 90 percent from 2019-2033” in AK and HI because if there isn’t enough done to make IFR happen all over AK, then that is just a pie in the sky wish that is meaningless to the extreme.)

So, yes Alaska is a problem and everyone knows how to fix it. Spend the money there or expect this issue to remain.

Other than medevac and AK, one thing I like to look at is other types of operation as far as cargo, on demand passenger or commuter. Ameriflight had a cargo accident last year as did its sister company (same owners), Wiggins Airways. The Wiggins crash was a double fatality training accident while the Ameriflight, which resulted in a serious injury, is notable because it flew as single pilot crew in a Beech 99 and the pilot only had 26 hours in the make/model which, for me raises all sorts of questions. The cause there was loss of control on takeoff. (Wiggins crashed again in January 2024 with a serious injury to the pilot — that looks like some sort of situation with the pilot door, but no final report yet.)

What I’ve found over several years of doing this is that you really need a five year survey to consider bigger trends and I’m not there yet in U.S. Part 135 stats. I can tell you that there are sightseeing accidents, helicopter with external load accidents (these are under Part 133 but if they involve a Part 135 operator, I include them), and lots of positioning flights classified as Part 91. Basically, it’s all in there.

The aircraft themselves range from single engine Cessna (206 or 207) up to Lear jets. And while Alaska is the primary state, they do happen everywhere and often the worst crashes, in terms of death or injury, are not in AK.

In 2023 there were six people seriously injured on Papillion Airways (sightseeing) in Nevada, and also the five fatality Guardian flight crash in Nevada, four deaths for Maritime Helicopters off the northern coast of Alaska and three deaths off San Clemente, CA for a Lear 35 with Phoenix Air Group. The worst accident, by far, in the past couple of years was 10 people killed on Friday Harbor Seaplanes off the coast of Washington in a mechanical issue with the De Havilland Otter.

What I am looking for

Part of why I do this is to see what is missing. In the cargo accidents, I want to know who they are flying for and generally, that information is not included in the accident reports. I want this to see if there are more issues with one customer - FedEx, UPS, DHL, etc. The NTSB does not collect this data however and I’ve had to go hunting for it on my own in news reports. After researching Key Lime Air flight 308 however, I really think the customer demands/scheduling could matter a lot in cargo investigations.

I’m also looking for types of operation, (hence the article I did for AIN on cargo companies and the current research I’m doing on medevac), repeat offenders and regions where the companies are based to see if there is a FAA oversight issue. (This was the case in Southeast Alaska as I found while researching the Taquan Air Service accidents in 2018 and 2019 for AIN.) While I do track aircraft type, I am unsure what I’ve learned there yet and collecting flight time is still important, but I’m comfortable saying it is not an overall predictor of pilot behavior.

The Major Point

Now, we come to the probable causes and that is the biggest deal for me.

First, I don’t worry much about things like bird strike, ground collision (when not caused by the pilot) or weird incidents like when ground crew did not stay clear of an aircraft, or medevac crew had a lithium battery blow up in their flight suit (OUCH) or when a boat crossed in front of a floatplane. (It happens.) There’s also things like an aircraft encountering a sinkhole while taxiing that I really don’t think point to any larger issue in the industry. (Not that any of these things aren’t important or terrible, but rather that they aren’t necessarily about the specific companies involved or FAA oversight.)

(I had to include this picture - LOOK AT THAT SINKHOLE!)

But there are lots of other causes such as fuel starvation, loss of control on landing or takeoff, unsuitable selection of landing site, failure to maintain clearance, failure to maintain altitude or stuff like filling out your logbook and flying into terrain because you weren’t paying attention, or taking off without realizing one of your engines was not working or getting out of the airplane while it was running so that it taxis - by itself - into a hangar, or improper braking action, or VFR into IMC, or continuation of an unstable approach which all, to one degree or another, point to issues with training, decision-making or operational control. (I bring up operational control because I wonder why certain pilots were in certain situations and the management policies/choices that put them there.)

It is in analyzing all of this that I learn about issues in Part 135. Part of the problem though is that if you read accident reports, a lot of the time the cause starts and stops with the flight crew and the interviews in the docket are sparse, if they are included at all. For me, that’s not enough but it is rare — and generally only when it is a high profile and/or a multiple fatality crash — that the NTSB goes beyond the pilot to find answers. And if you are hoping for interviews with the FAA Certificate Mgmt Team to see what they thought of a company’s operations, well, you are really hoping for a lot.

I have read thousands of accident reports. THOUSANDS. I can do this because most are less than ten pages long. As an example, let’s talk about Hageland Aviation.

Time for an example from Alaska

In May 2013, there was a VFR into adverse weather conditions accident in Newtok with Hageland Aviation, one of the largest Part 135 commuters in the U.S. The pilot was flying a Cessna 207; there were four minor injuries. Here is the entire analysis of the crash from the accident report:

And here is the only interview included in the docket (I have no idea why the referenced passenger’s statement is not included). Mr. Griffin was one of the FAA inspectors; I do not know if he was assigned to the company.

Remember - one of the largest Part 135 commuters in the U.S. flying 252,687 scheduled passengers that year but this is all you get in the investigation.

In November 2013 Hageland had a VFR into IMC crash with a Beech 1900 in Badami; the crew and single passenger had no injuries. The co-pilot submitted a one page statement, there is a one page summary of the interview with the captain and then there is this one page summary of the interview with the station manager/flight coordinator:

Be aware that for Hageland, Station Managers had operational control. You can see from this statement however, how little that meant to flight safety.

One week after the BE1900 accident there was a VFR into IMC crash in a Cessna Caravan that killed four and seriously injured six in St Marys. This accident made national news but even in its docket, which is about 200 pages total, there is only a four page interview summary with the company POI. Only four pages even though he tells investigators this:

The other failures referenced here include events and occurrences that did not meet the classification of “accident” but were ongoing prior to the St Marys crash.

The next Hageland docket, for a double fatality training accident in April 2014, had basically no interviews. At that point the FAA began a formal audit of the company, the NTSB issued urgent safety recs, it was a whole big deal. But they kept flying. They also kept crashing and it wasn’t until its parent company, Ravn Air Group’s bankruptcy in 2020, that Hageland Aviation went out of business.

In all of that, and many other accident reports for Hageland, I looked for interviews with all of the company’s POIs. When I say “all the POIs”, I do mean all of them, which according to the interview with the FAA Frontline manager after St Marys was a sizeable group:

From the time period that I started 2½ years ago, when I got there it was Corey Howlett and Corey had been in that position for some time; I don't know exactly how long. But he got a transfer to the UPS certificate, so his replacement was a -- guess it was just an assignment that someone had to do, temporary assignment until we could pick somebody. I believe that was Matt Waln and then we had a bid put out and Angelique Talbot got that bid. It wasn't, maybe 6 months later she got a job down here at the region and so Danny Larson took over again on a detail…And then Scott Schweizer again on a detail and then finally I was allowed to bid out the job on a more permanent basis and again Danny Larson became the POI.

So, in 2½ years there were six changes of operations inspector through five individuals but you wouldn’t know that unless you read this one specific interview for this one accident. Otherwise, these folks are not really mentioned in the other accidents and rarely talked to. But if you look at what happened with this company, it’s clear the shifting FAA oversight was a huge problem.

This all occurred in Alaska, so I know a lot about it. When I look at accidents in the entire U.S., I’m always wondering what was going on with those companies and where I have to look to find those answers. I do think the answers are worth looking for however, just as I think that Part 135 deserves a lot more attention from industry publications. There are many ways to improve the safety record; we just need the specific data to do it.

Somehow, I think I have ended up (almost by default) as a fulltime Part 135 journalist.

A Housekeeping Note…

I’m going to try and get an aviation-specific newsletter out every other week and mix them up (in the intervening weeks) with something that is a bit less research-heavy as these take me so much more time to put together. Next time will likely be about some writing I’m doing, specifically of the Cosmic Ray Expedition kind, because I’ve been thinking about something there I’d like to share. Anyway, trying to get regular on all this substack suff, so you know what to expect, and trying to stay interesting so you continue to read and subscribe and, hopefully, tell others about me. Newsletter are tricky and I’m still finding my way, and I thank you for patience as I get this in the best possible shape for readers.

Gratuitous airplane picture…this is unloading the mail (bypass) in Scammon Bay in the ‘90s.