Mid-air Collision of Wingsuit Flyer and Pilatus PC-6
On the 27th of July 2018, an experienced pilot departed with ten parachutists on board for a routine drop mission. It was the pilot’s fourth flight of the day in the Pilatus PC-6 Porter, a single-engine aircraft popular with parachuting and skydiving operations for its short take-off and landing capabilities and low maintenance requirements. The flights were operated by the École de parachutisme de Midi-Pyrénées (Midi-Pyrénées Parachuting School) at Aérodrome de Bouloc in Southern France.
The pilot of the accident flight was the chief pilot of the parachuting school and held a commercial pilot license, a class instructor rating and a Pilatus PC-6 rating. He had a total of 13,366 flight hours, including 7,560 hours on the PC-6. His Class 1 medical certificate mandated a second qualified pilot for all commercial flights but he later said that he believed that he was legal to fly solo.
The weather was clear and warm that day, with a wind of 5 knots on the ground and 15 knots at 4,000 metres, the final drop altitude. The deputy technical director of the parachuting school gave a field briefing to the parachutists. The briefing was focused on the parachute landing options, including alternate landing areas (in case the jumpers could not reach the aerodrome) and the precautions necessary if they found themselves near power lines. It did not include the manoeuvre zones or the aircraft descent zones.
The pilot did not give any further briefing to the parachuters.
Two of the parachutists wore wingsuits, a type of jumpsuit with webbing between the arms and the thighs to act as an airfoil. The increased lift allows the wearer to glide before deploying the parachute to slow the descent for landing.
This two-minute video shows a straight jump from a rocky outcrop through a cliff (if you are reading this in email, you may need to click through to view the video).
The top two comments are:
Lets you appreciate the huge tasks birds do just to say alive.
and
I get the feeling there are no small mistakes in this sport.
Aircraft jumps are considered to be much less dangerous.
The BEA accident report from the jump at Bouloc refers to Paralog Performance data for precise data for the wingsuits worn by these two flyers.
As soon as they exit the aircraft, wingsuiters convert part of their vertical speed into horizontal speed. This phase lasts about fifteen seconds, during which the wingsuiter’s lift-to-drag ratio increases from about 0.5 to 2.5. The lift-to-drag ratio then stabilizes between 2 and 2.5.
The first parachutist was to jump at 1,500 metres (~5,000 feet) and then the other nine at 4,000 metres (~13,000 feet). The two wingsuit flyers would be dropped last.
The flight departed from runway 28 at 10:00. Although the report doesn’t mention it, the first parachutist presumably jumped normally at 1,500 meters. The PC-6 continued to climb to 4,000 metres, where the next seven parachutists jumped from the plane at six second intervals.
The French Parachuting Federation specifies that when a wingsuit flyer exits the aircraft, he or she should then continue in a path parallel to the aircraft for twenty seconds. There’s no procedure for the pilot, although the pilot would generally initiate a slight descent and negative attitude to keep the horizontal stabiliser as far from the parachutists as possible. Investigators found that after a drop, the PC-6 generally descended at a vertical speed between 3,500 and 5,500 feet per minute, making for a slope of 20-30°. The wingsuit flyer’s slope would be at 10° after the jump and into the glide parallel to the aircraft. After 15 seconds of descent the slope increases to 35° and then drops to 20° in stabilised flight.
That day, the pilot slowed the PC-6 to 65 knots. The first wingsuit flyer exited the aircraft and continued on a path parallel to the aircraft. A few seconds later, the second wingsuit flyer followed. The wingsuit flyers disappeared out of his view, which was normal. The pilot started his descent back to the airfield. He didn’t have visibility of the flyers but entered a left turn to avoid flying into their path.
The second wingsuit flyer had a camera mounted to his helmet. The camera filmed the first wingsuit flyer in front of him. Suddenly, the PC-6 appeared a few metres in front of him.
The aircraft shook with a violent impact. The aircraft’s left wing had clipped and killed the wingsuit flyer in front. The wingsuit flyer’s parachute deployed and the body drifted down to a field.
The pilot was immediately aware that he must have collided with one of the flyers. The left flap was damaged, but the PC-6 was otherwise flyable and the pilot returned to the aerodrome. The pilot believed that he had descended normally and that the wingsuit flyer must have been off-course.
Investigators used the video from the second wingsuit flyer and the aircraft’s navigation system data to recreate the seconds before impact.
The first wingsuit flyer jumped at 4,400 metres, specifically 14,439 feet, and then glided parallel to the aircraft. The pilot said that his vertical descent speed was normal for a descent after a drop, estimating it as between 3,500 and 4,500 feet per minute. However, the data showed that this was not the case. The PC-6 started to descend even as the second wingsuit flyer jumped, flying a downward slope of about 50°, instead of the expected 20-30°. The PC-6’s average vertical speed was 5,800 feet per minute which increased to an average of 6,700 feet per minute from the time that the second wingsuit flyer had exited the aircraft until the collision. The wingsuit flyers could not see the aircraft, which was (or should have been) above and behind them. It would be difficult at best for the pilot to see the wingsuit flyers.
The first wingsuit flyer had been in the air for eleven seconds, still following the path parallel to the PC-6, when they collided at 13,400 feet.
The following graph shows the expected descent of the Pilatus PC-6 and the immediate descent that led to the collision with the wingsuit flyer.
There is no specific procedure for a post-drop descent. The French Parachuting Federation 2016 safety report regarding mid-air collision focuses on the risk of another aircraft after the parachute is extended. The issue of a wingsuit flyer and aircraft colliding during the first seconds of the jump does not appear to have come up as an issue previously. Wikipedia has a List of fatalities due to wingsuit flying; almost all of them were BASE jumps (from high ground, for example cliffs, bridges or towers) and involved crashing into terrain. In fact, the linked article specifically clarifies that “aircraft descending a much less-deadly form of wingsuit flying.”
The final report was released in 2020 and cites the following contributing factors:
- lack of an onboard briefing between the parachutists, the wingsuiters and
the pilot; the French Parachuting Federation (FFP) not identifying the risk presented by the
coexistence of aircraft and wingsuiter paths immediately after exiting the aircraft;the immediate descent of the Pilatus, on a steep slope, commanded by the pilot even though he did not have visual contact with the wingsuiter(s).
Two years after this report, the mayor of Chamonix (in the French Alps) banned wingsuit flying for six months, after eight deaths in Franch that year, of which five were in Chamonix.
His view was that the sport was not regulated enough and the flyers did lacked experience for the flights they were attempting. If you are interested in more detail, Matt Higgens wrote a very good article about the ban and the dangers of wingsuit flying in Outsider Online last year: The Most Dangerous Part About Wingsuiting Might Be the Wingsuit.
You can read the final report on the BEA site.
As a result of this tragic crash, the French Parachuting Federation published an information bulletin about the need to define and separate the aircraft’s route from the wingsuit fliers, followed by a technical directive.
Pilot briefings for jump flights are now mandatory. The parachuting school has changed their procedures after the drop, asking pilots to adjust their power and maintain level flight for 10-20 seconds to ensure vertical and horizontal separation.
Last month, LeParisien reported that the pilot was found guilty of involuntary manslaughter and breaking the conditions of his medical certificate by flying alone. He was given a 12-month suspended prison sentence and banned from flying for one year. The parachuting school was fined €20,000 for not having confirmed the state of his license and medical certificate.
When I was skydiving (and reading a lot about it), steep aircraft descents were common because the plane isn’t making any money when it’s descending; since a skydiver in standard suit and position falls ~7500 feet/minute (after a few seconds to get up to speed), there’s not much risk of collision if the plane waits a bit after the last skydiver exits before heading down, and if there are standards for landing (e.g., “The plane makes downwind over the tree-covered/water/… side of the runway while the canopies aim for the grassy side.”) they don’t run into each other near the ground. But paired with this is a rule I learned in scuba: “Plan the dive; dive the plan.” — which was also usually observed (if not enunciated) at the drop zone I went to. I’m not seeing a link to a report on this accident, so I can’t tell whether this pilot had previous experience with wingsuits and/or had talked at all with the two wingsuiters, so it’s not clear whether the pilot was inattentive or unprepared — although the incorrect claim about descent speed doesn’t say anything in his favor.
Oops, the final report was in there but lost in the text. I’ve added a clear link at the bottom but unfortunately, it doesn’t answer your question. I’m not sure it is relevant, though, as it doesn’t seem to have occurred to anyone how easily the aircraft could catch up with the wingsuit flyers.
I have a pic of the vertical stabilizer of out pc6 passing between my legs
eeep!
OK, this is embarrassing but I guess I’m still not recovered. Somehow I managed to post without the last three paragraphs of the article. I have now added them.
As a result of this tragic crash, the French Parachuting Federation published an information bulletin about the need to define and separate the aircraft’s route from the wingsuit fliers, followed by a technical directive.
Pilot briefings for jump flights are now mandatory. The parachuting school has changed their procedures after the drop, asking pilots to adjust their power and maintain level flight for 10-20 seconds to ensure vertical and horizontal separation.
Last month, LeParisien reported that the pilot was found guilty of involuntary manslaughter and breaking the conditions of his medical certificate by flying alone. He was given a 12-month suspended prison sentence and banned from flying for one year. The parachuting school was fined €20,000 for not having confirmed the state of his license and medical certificate.
Another note: I’ve never heard of a medical certificate requiring that the holder have a copilot. Is this something peculiar to France, or to your side of the pond generally, or has it also come up in the US? What’s the purpose — does it let someone with a periodic condition keep piloting? I have this vision of somebody known to occasionally get the shakes being a pilot, but that seems more like a comedy routine than a real-world case.
It’s called an “operational multi-crew limitation” (OML for short), and you can find the rationale for it in the ICAO “Manual of Civil Aviation Medicine”. Not all states implement it, but it does seem common in Europe. A typical trigger would be a heart condition. For example:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6019020/
In the wingsuit collision documented here, the pilot held the position of Chief Pilot at his company, which is probably why his restriction went under the radar.
It’s good to see Sylvia back at the helm!
Interesting; I hadn’t realized there was a legally intermediate area between being approved and being downchecked. (I don’t know whether this is because I didn’t have to know this, having had only a basic certificate and stopping flying in my 20’s, or because the rules were changed due to pilot pressure and more information about activity after cardiac issues/treatment.) I note that the article is incorrect about levels of certification; https://www.ecfr.gov/current/title-14/chapter-I/subchapter-D/part-67?toc=1 confirms that the US still has 3 levels, ~corresponding to the 3 levels of license (transport, commercial, recreational) described under https://www.faa.gov/faq?combine=&field_faq_category_target_id=11581 (search for “class” or “student” and click the arrow);
EASA has
• class 1 medical certificate for commercial pilots
• class 2 medical certificate for private pilots
• class 3 medical certificate for air traffic controllers.
See https://www.easa.europa.eu/en/domains/aircrew-and-medical/medical .
(The class 1 certificate encompasses class 2.)
The issue here is separation (as it obviously is with all mid-air collisions). The BEA report notes that the pilot stated he “thought that he was deviating from the supposed path of the wingsuiters by taking a left turn”, but the flight data shows that “the pilot initiated a left turn after the collision, while still descending”. The immediate steep descent removed the vertical separation, and the pilot’s failure to turn before the descent meant a lack of horizontal separation.
It would probably be safest to have a procedure like “wingsuits break right, pilot breaks left”, but I’m not sure if the wingsuit is controllable in that phase of flight?
I wouldn’t expect a wingsuit to be seriously controllable in its first few seconds; the links above speak of it being pushed into ~airfoil shape by the air it’s falling through (just like a standard modern canopy after it comes out of the container). I’d also guess wingsuit flyers would prefer to make a clean exit rather than having to steer immediately. If the airplane just keeps moving forward at altitude, drag would hold back the flyer, giving both horizontal and vertical separation; possibly breaking in opposite directions after several seconds have passed would add to the safety that comes from the plane not diving immediately.