On the surface, the two pilots could not have been more different.
One was an ATP-rated professional pilot. He was 48 years old, had a first class medical, a handful of instructor ratings, a Boeing 737 type rating, and a logbook heavy with nearly 8,000 hours.
The other, also male, was a 71-year-old private single-engine land and glider pilot on BasicMed, with a lifetime total of around 1,500 hours.
The pro was flying a dashing red and white Pitts Special S1S, while the older gent was flying a beautiful antique black-and-orange Waco that was a decade older than he was.
Except, of course, technically, neither pilot was flying at the time of the accident.
But despite their differences, the NTSB found them to be more alike than they first appear — and equally to blame for what happened.
The Accident
It all went south for the two pilots early on a sunny Saturday afternoon in August 2022. It had been a busy day at the Van Sant Airport (9N1) in Erwinna, Pennsylvania, north of Philly. There had been vintage airplane rides, school flights, glider tows, and visiting pilots flying in and out.
I bet there were even free hot dogs involved, but none of the official documents mention that.
The airport has a pair of turf runways: 5G/23G, at 200 feet wide by 1,340 long; and 7/25, at 120 feet wide by 3,058 long, the ends charmingly marked with white-painted tires. The shorter runway intersects the longer one at about the midpoint of the longer runway, with the angles close enough that a diagram of the airport, if an official one actually existed, would resemble a partly submerged cartoon alligator with its top jaw open wide.
The Waco pilot was based at a nearby airport, and had flown into Van Sant to collect on a debt: One of his buddies “owed” him a Stearman ride.
After a pleasant morning visiting his airport friends and collecting on his debt, the Waco pilot prepared to depart, and taxied out toward Runway 25, which he knew to be the published “powered” airplane runway. Unlike many airplanes both based on, or visiting, the field, the Waco actually had a radio. The pilot even had it turned on, but he made no taxi calls as he headed out.
As fate would have it, slightly before, the Pitts pilot, who was based at Van Sant, was also taxiing out for a flight, but he decided to takeoff from the glider runway, as the locals often did. It was the smoother of the two runways.
And as fate would further have it, the smooth glider runway also has a significant hump in it, so that a pilot in an airplane on the threshold can’t see the far end. Said another way, the pilot of the Pitts, as he made his radio call and throttled up, could not see the Waco, which was at that very second crossing the end of the glider runway to get to the powered aircraft runway.
The Waco pilot, being a frequent visitor to Van Sant over the years, knew that you couldn’t see the end of the glider runway from the crossing point, but he also knew that if a tow plane was in place, it would be visible, as the tow lines were sufficiently long to place a tow plane far enough downfield to be visible. He didn’t, apparently, visit often enough to know that the restriction on powered aircraft using the glider runway was lightly, if ever, enforced at the airport.
All looking clear to the Waco driver, he began a leisurely taxi across the wide strip of grass. He even heard the Pitts pilot make his takeoff announcement on the CTAF, but at that second the older pilot suffered a common phenomenon called expectation bias.
Even though he heard the on-the-go call for Runway 23, his brain processed it as Runway 25 — as he expected (hence expectation bias) the Pitts to be using the powered airplane runway. He was actually looking down Runway 25 and wondering why he couldn’t see the Pitts when it blind-sided him. Literally.
Not quite truly airborne yet — the accident would be classified as a ground collision — the departing Pitts hit the Waco, ripping its top right wing clean off. The Pitts pilot never saw the Waco and the Pitts — traveling faster and semi airborne in ground effect — fared worse out of the deal, ending up a sad crumbled little ball of red and white racing stripes.
Luckily, neither pilot was seriously injured.
The NTSB
The NTSB investigators must have flipped a coin on this one, but I guess it landed on an end, so both pilots were thrown under the bus.
The NTSB’s final report lists the probable cause of the accident to be: “The inability of both pilots to see and avoid the collision. Contributing to the accident was the pilot of the red airplane’s decision to utilize the converging runway for takeoff, contrary to the guidance published in the airport facility directory. Also contributing was the failure of the pilot of the black airplane to adequately utilize his radio to announce his taxi on the common traffic advisory frequency and his failure to recognize that an airplane was departing from a converging runway.”
Is that a fair analysis? Did it really take two to tango or was one pilot more at fault than the other?
Analysis & Discussion
So what do we have? We have a less than ideal airport layout with intersecting runways, view-limiting terrain, and published data ignored by the locals. Why, yes, that does seem like a good recipe for an accident, doesn’t it?
But let’s consider the degree of blame in this dance, starting with the Waco pilot.
The NTSB singled him out for failing “to adequately utilize his radio to announce his taxi.” What about that?
AIM 4-1-9 does recommend making a call “before taxiing” at non-towered airports, although that information is buried in a table, and is not in the main text.
Meanwhile Advisory Circular 90-66C Section 10 — which the FAA says “is intended to supplement the AIM” on the subject of communications at non-towered airports — says that taxi traffic should “monitor/communicate” on the CTAF.
The Waco pilot was monitoring, but not communicating, which might actually be appropriate because, if you read on further, the AC cautions pilots to “limit communications on CTAF frequencies to safety-essential information regarding arrivals, departures, traffic flow, takeoffs, and landings.”
There’s nothing in there about taxi.
Personally, I think that most taxi calls at non-towered airports — in general — are a waste of everyone’s time. There is one exception, however, and that’s when crossing a runway. Any runway. Hell, at a towered airport you need express clearance to cross closed runways, even ones that have been permanently closed. So I’ll give the Waco driver a pass on not making a taxi call, but he should absolutely have made a runway-crossing call.
As to the Pitts guy taking off from the glider field, who received scorn from the NTSB for his “decision to utilize the converging runway for takeoff, contrary to the guidance published in the airport facility directory,” what percent of the blame does he legitimately earn?
A couple of things come into play there.
In his communication with investigators, the Pitts pilot reported that shortly before he headed out, another airplane took off from the glider runway. So in one respect, he was using the “active.”
He also served as a glider tow pilot at Van Sant, so he might have been “used to” using that runway.
And of course, we know it was favored locally for being smoother.
But at the same time, as a local, he knew about the visibility issues around the hump, and as a glider tow pilot he would normally have taken off from further down the field. One would think that an 1,100 pound airplane with a frickin’ 200-hp IO-360, at that altitude, could have spared the 200 or so feet of runway that would have improved visibility.
That said, no one seems to be asking why such a fire-breathing little beast was only a few feet off the ground at the time of the collision, at a location where he was pretty much running out of runway.
So maybe the coin really does land on end.
But aside from assigning blame to the two pilots, there’s also some blame to be found in the larger way in which things were run at the airport at the time, which I would define as an issue of local culture. A local pilot culture that was in the habit of using powered airplanes on a G-designated runway.
That’s a problem on several levels — the largest of which is that visiting pilots don’t know what the local social norms are.
It seems to me that if you establish any sort of local standard operating practices (SOPs), regardless of how they fit into the normally accepted external SOPs, they need to be published. And, in fact, in defense of the airport, it pretty much has been since.
The Chart Supplement at the time of the accident advised pilots that taking off and landing on the 05 side of the glider runway was prohibited, and that the 23G side was for gliders only. As we now know, this was lightly enforced — if at all — at the time.
But now, perhaps because of the accident, the entry for the airport in the Chart Supplement says that Runway 23G is for “based aircraft use only.”
So the local norms haven’t changed, but at least outsiders get a head’s up.
After the crash, the Pitts pilot lobbied for the creation of an official hot spot, and even submitted a proposed diagram to the NTSB (below).
There are two problems with that suggestion. The first is that it’s the FAA that designates hot spots, not the NTSB. Plus the airport does not have an airport diagram, so there’s nowhere for the hot spot to be posted. (The airport’s web page now has a similar diagram, but visiting an airport’s website isn’t a normal part of preflight planning for most pilots.)
The Takeaways
So I guess the main takeaways are — if you have a radio — always, always, always make a radio call when crossing a runway. And tied in with that, no leisurely crossings. Lastly, when crossing a runway, even at the end of one, but especially at a wide one like that one, the major focus of your attention should be down the runway rather than forward.
Oh yes, and be on guard against expectation bias. We need to be most careful to hear what is actually said, not what we expect to be said.
And, perhaps, as it is our regulatory requirement to get “all available” information when planning a flight, that in addition to checking all official sources, we should pay a visit to the website of any new airport we are dropping in on — just in case the locals have shared some tips on their culture.
The Numbers
Want to read more? Download the NTSB’s final report here or view the items on docket here.
Uncontrolled Airport. In Australia we have 100’s of them so it is pilot beware, see or be seen. Even at my home base YWVA, this can be a very busy airport and still classified as an ALA. It also happens to be in a region where GA aircraft are transiting to other Airports/ALA’s (Uncontrolled) within the CTAF area. Our protocols are that our Taxiing call also includes the designated runway we are intending to use and flight profile (Circuits, Local Flight, Airwork or intended destination). Like some ALA’s we do have a preferred runway even with nil wind. However, with traffic in the pattern they are all calling the designated runway on their downwind/base calls so any itinerant traffic approaching the airfield knows what runway is in use. This works extremely well and creates excellent pilot situational awareness even when we hear traffic calls from another nearby aerodrome. The last time we had an incident which resulted in a collision was within the last twelve months in Queensland at an ALA was due to one of the aircraft not listening out, was not aware of the other aircraft on the cross strip. It is still to be determined whether the offending aircraft didn’t have their radio switched on, or if on, their volume set at an inappropriate level, or their equipment was u/s. Short answer, communication by way of radio calls is essential at any uncontrolled aerodrome is my view and opinion.
If there’s a real, logical reason for a taxi call, do it. But the vast majority of taxi calls are overkill. The one I hear most often in that category is “Greeley traffic, xxx is taxiing from the ramp” with nothing more.
I always make a taxi call crossing any runway, but only rarely make any others except to announce that I’m either departing or clearing from a runway.
Incidentally, at Greeley (KGXY), the published calm wind runway is 35, but almost everyone, both locals and transients, use 10 when the wind is calm, because it’s more convenient to the ramp. Except of course the heavier, faster ones typically use 35 for its length.
Mr. Dubois—to be concise, too much information is better than none at all.
Make radio calls for ALL movements whether on the ground or in the air including in the environment of uncontrolled airports. Turn radio calls into a habit rather than an optional procedure.
First I’ve seen this report. Stopped reading at big get together, turf runways, possibly 🌭.
I would only reiterate again, though usually the first question out of someone’s mouth…hours quite often don’t really matter.
Ouch, I hate it when that happens!
Have experienced a ‘mid-air’ myself; except in my case, the tow plane (me) was hit from behind, while still on the ground, by an airborne Cirrus sailplane. My tow plane, a PA-25 Pawnee, was barely ‘scratched’. The very-expensive, high-performance Cirrus, was “substantially damaged”….
As far as this incident goes…
“Except, of course, technically, neither pilot was flying at the time of the accident.“
Technically: They were.
At least the FAA & ICAO think so.
The definition of “flight time” is found in 14 CFR 1.1, General Definitions:
“Flight Time Means: Pilot time that commences when an aircraft moves under its own power for the purpose of flight and ends when the aircraft comes to rest after landing.”
That includes ‘taxi time’…assuming they intended to take off (which apparently, the Pitts had…). See the 2004 “Kania” FAA interpretation.
It may have been classified as a “ground collision”, but…technically…they were flying.
Interesting read, quite a chain of events leading up to the accident! Breaking any link in the chain would most likely had a different outcome.
I thought the comment on the Waco pilot’s “leisure” taxi was amusing as I own a 1929 Fleet and leisurely would probably describe my taxi method. The Fleet has no forward visibility on the ground (S-turns are mandatory) and the brakes are just barely powerful enough to be called brakes.
That being said, I have to ask…. why was the 200 hp Pitts only a few feet in the air at the end of a 1340 ft runway?
If power pilots are using the glider runway because it’s smoother, maybe one way to prevent this kind of thing in the future would be to fix the power runway so it’s as smooth or smoother. That removes the motivation and temptation to use the inappropriate runway.
Responding to: “So what do we have? We have a less than ideal airport layout with intersecting runways, view-limiting terrain, and published data ignored by the locals. Why, yes, that does seem like a good recipe for an accident, doesn’t it?”
Very detailed analysis above, but belabors the issue which breaks down to two simple observations: Waco should have used his radio, and Pitt should have used the correct runway for his takeoff. Had they both done that, there wouldn’t be two damaged airplanes.
The root cause of this mess is the all-too-human willingness to ass-u-me (makes an ass out of you and me) a set of circumstances rather than verify actual conditions. No excuse for Waco not keying the radio and announcing his taxi intentions. No excuse for Pitt just following the lead of the other takeoffs making unauthorized use of the glider runway. By the Grace of God they both lived, but it was mighty close to funerals that day. Moral of the story: in GA and most other walks of life, don’t assume anything. Always check it out.
Regards/J
I disagree with a statement made by Mr. Dubois in the article. The statement reads, “ Personally, I think that most taxi calls at non-towered airports — in general — are a waste of everyone’s time. There is one exception, however, and that’s when crossing a runway.”
There are times when the frequency is quiet and it might appear that there is no traffic. A taxi call, either by myself while taxiing or by another aircraft while I’m airborne has proven quite helpful many times.
For example, let’s assume a non-towered field with a single runway. I usually start monitoring CTAF 15-20 miles out. If I hear another aircraft make a taxi call prior to my first call, usually at 10 miles, I will adjust my pattern to use the same runway the taxiing aircraft intends to use, conditions permitting. Likewise, there have been times where my taxi call is the first and only call an inbound aircraft has heard. My taxi call has allowed the inbound aircraft to plan accordingly. Perhaps I didn’t get on frequency until after a jet has made a call and is established on final. My call has prompted the jet driver to make another call increasing my situational awareness.
Our airport has a very interesting mix of trainers, experimentals, helicopters, jets, warbirds, and other GA traffic. I have found that when the frequency is quiet, it is often assumed there isn’t any traffic. This is when a taxi call is critical. That said, we don’t need 20 flight school airplanes making a taxi call every two minutes. While there are times when a taxi call can be distracting, there are times I welcome it.
Jerry Esquenazi
I make a taxi call prior to moving from the parking space on the north side of the hangar. My initial taxi will be eastbound and parallel to the north side of the hangar. I want to know if someone if taxiing northbound along the east side of the hangar. I’d rather not be surprised by the sudden appearance of another airplane at the northeast corner of the hangar.