#20: Only a few seconds to make the right decision

We don’t talk enough about timing in aviation - the few seconds to make the right decision when something unexpected occurs and the pilots can not be wrong.

JeJu flight 2216 wreckage. Photo credit Hyung-Jin Kim, Associated Press

I completed two paying assignments over the holiday break and I’m working on more. I planned to write about the USPS and aviation in Alaska for this newsletter as it could become a topic of significant concern for the industry (and the state) in light of the incoming administration’s threats to slash the federal budget. But then Jeju Air flight 2216 crashed killing nearly everyone onboard and a lot of folks with a lot of opinions started sharing them. I have some thoughts as well, but they aren’t about what we don’t know. FYI, what we don’t know is pretty much everything.

When a high profile aviation accident occurs, the obvious information is presented very rapidly. In the U.S. a NTSB GO Team will be dispatched, often including a boardmember, and there are regular press conferences and releases to the media. As someone who writes primarily about smaller crashes, I am more accustomed to the long wait to find out anything of substance. Right now I have dozens of accidents in my Part 135 database from 2023 and 2024 that have only preliminary information; a not insignificant number lack even that. The oldest accident without a final report is from July 2022. (I have no idea what is going on with this accident - there was a documented autopilot issue in a Challenger at 20,000 feet after departing from San Francisco and a flight attendant sustained a broken ankle. The aircraft returned to San Francisco and landed without issue. Why it has taken over 30 months to issue a probable cause in this crash is a mystery.)

Analysis of flight 2216 will move far more rapidly than the 2022 Challenger accident, but still we need to wait. In the meantime, with some puzzle pieces for the accident already revealed, (a possible bird strike, a missed approach, a “mayday” declaration, landing gear issues, a possible engine issue), the thing I can’t stop thinking about is how little time all pilots have to make the right decision in a moment of crisis. The wrong first decision, in those few seconds in which you must decide, can lead to a cascade of increasing damage and destruction; it can make survival impossible.

The Crashes I Still Think About

When I was just out of college two friends of mine, both pilots and fixed wing instructors, were in a R22 helicopter when the main rotorblade diverged for unknown reasons, resulting in mast bumping and contact with the cockpit. In other words the rotor blade dropped down through the cockpit. My friends were injured and likely killed in the air. The one who was flying, and apparently serving as instructor, had almost 3,000 hours of flight time in fixed wing, and only 9 hours in helicopters. He flew an hour earlier in the R22 with another former classmate who was also a fixed wing instructor; she told investigators they practiced low rotor RPM recovery. What we think ja[[emed is he was doing that again and did not recover properly. The recovery for a rotorcraft at low RPM is exactly opposite that of a fixed wing stall. So, my friend (one of the best people I have ever known), made a mistake by reacting as if he was in an airplane in his first action when he encountered trouble at low RPM and the blade impacted (“bumped”) the mast, causing it to drop further and contact the cockpit. In seconds, it was over.

In Alaska, I had a friend who was chasing wolves, flying a cargo flight at low level in mountainous terrain. He followed the pack into an unfamiliar canyon and then waited too long to start his climb. He impacted the ridgeline fifty feet from the top; he did not survive. Another friend, and former co-worker, also on a cargo flight had some sort of undefined engine trouble in a twin engine Piper Navajo on takeoff from a small airport in Interior Alaska. He told the local Flight Service Station he needed to return for landing. He came back around but did not have the necessary altitude or airspeed to make the runway; he impacted the Yukon River and was killed. A portion of that aircraft was not recovered and it remains an undetermined accident. (I wrote about these two accidents in my book.)

I mention these three crashes to illustrate a point about the time to make the right decision. You have to do what is right immediately and if you don’t, you often do not have that opportunity again. My friend did not react correctly in the helicopter; my friend waited too long to climb out of the canyon while looking at the wolves; my friend should not have tried for the airport but instead attempted landing on a gravel bar. They did not make the right first decisions and so they did not survive.

When Pilots [Sometimes Famously] Get it Right

We often talk about US Air flight 1549, the “Miracle on the Hudson” as a perfect example of everything working as it should: right crew, right place, perfect weather. The landing was impeccable and a tribute to Capt Sullenberger and First Officer Skiles. But that landing, as great as it was, was not the most important thing about the accident. It was the captain’s quick decision to land on the Hudson River, when they still had time to set up for the landing, that I think matters most. If they committed to LaGuardia or Teterboro and couldn’t make it, as the simulations later showed they would not, well, it would have been a horrific tragedy. The right decision when they still had time to make it was enormously significant to the landing’s success. (This is from the movie, “Sully” but it shows how quickly the decision was made.)

Years ago, when he was flying Part 135, my husband lost an engine on takeoff in Fairbanks in a Piper Navajo. It was dark (early morning), in a snowstorm, and he was heavy - full fuel and cargo. He turned into the good engine, lined up on the roads where he had some light, told the tower he needed immediate clearance to land¹ and he brought the plane around. It all went like clockwork and in retrospect, might seem easy. That would be an incorrect assessment, however. The first decision that set up everything else was when he decided to fly the airplane and not land immediately. When he lost the engine he was still over the runway and only about 200 feet. But the gear was up and he couldn’t be sure he had time to lower it with enough runway to land. So he went around. He never gained altitude and his airspeed dropped - the stall warning was going off almost the entire time. He ended up landing on the ski strip because he couldn’t turn tight enough on final to make the main runway. But the landing was fine and other than the engine, the aircraft was undamaged.

(In case you were wondering, it was around this time that I fell in love with him. I was the AM dispatcher and manning the company radio when all this went down. Funny enough, I never thought he wouldn’t make it back okay and when he called to tell me he had to return, he was completely calm.)

If he tried for the runway, he would likely not have had the gear down and locked, he would have crashed, slid, and he was full of fuel. It would have been bad. It was the decision to land in the best place you can and not the perfect place, just as they did with flight 1549, that made the difference that day.

There are so many accidents I have studied where you can see, (from the comfort of the desk), that the right decision was not made first. In Key Lime Air flight 308, it was the decision to continue too long in bad weather; in the Guardian Flight Hawaii medevac, it was the decision not to tell the tower he was in trouble when the autopilot was lost (this likely would have cost him his job, but the tower’s help might have gotten the plane down safely). There were other decisions, long before these and made by others, that set those flights up for failure, but the wrong decisions in the cockpit were the ones that sealed it.

PenAir flight 3296

The NTSB found that PenAir flight 3296 overran the runway in 2019 in Dutch Harbor, AK, killing a passenger (this was a Part 121 flight) due to “the landing gear manufacturer’s incorrect wiring of the wheel speed transducer harnesses on the left MLG during overhaul.” I do not agree that this was the probable cause. The mis-wiring was done more than a year previously and the aircraft was flown continuously after that; it even had a flight earlier the day of the accident to the same airport.² The difference between the accident and all the other flights was the crew’s decision-making. They had extremely low flight time in the Saab 2000 for operating in Dutch Harbor; their time was actually a violation of the company’s own Ops Manual requirements for the airport³. (Dutch is an notoriously tricky airport.) When they landed, the crew was exceeding the aircraft’s tailwind allowance for Dutch at their weight. On final, the co-pilot told the captain the current winds and asked him if they should divert (it was their second attempt at landing). The captain refused. That was the wrong decision moment and everything that happened afterwards, the skidding and the runway overrun, was, I believe, because of that decision. (I wrote about this accident for the Anchorage Daily News before the final report came out and provided a lot of details about the aircraft load, etc. I will write about it much more hopefully one day.)

The passenger was killed when the prop impacted the fuselage; you can see that location in the picture above.

Another example: when the pilot of FedEx flight 8312, a Cessna Caravan landing in Wisconsin in 2020, could not see the runway while on final for a GPS approach, yet decided to land instead of going around, he also made a split second wrong decision. He ended up landing off runway and getting seriously injured⁴. You can read the entire NTSB analysis here - this is it, no docket, nothing else other than basic pilot, aircraft and weather info. I would love to know what the company (CSA Air) or FAA or even the pilot had to say about that decision but apparently, no one thought it was worthwhile to ask about.

Back to the Present

So, JeJu 2216. It could be the problems piled up too fast for the crew to respond; it could be there was a conflict in the cockpit within the crew that slowed their reaction; it could be there was a catastrophic problem we don’t know about yet that made any good response impossible; it could be that the wrong decision was made first and then there was no time to recover from it. We likely will not know what happened until the Flight Data Recorder is analyzed (it’s with the NTSB in Washington now). I hope it tells us a lot; I hope it tells us everything. Until then, I keep thinking about how little time you have when an airplane is in trouble and how no one can know what that is like (trust me) until they are in a similar situation.

Finally, there were 208 seconds for US Air flight 1549 from bird strike and dual engine failure to landing on the Hudson; you probably spent more time than that reading this newsletter.

1. They knew him at ATC - he was the earliest scheduled flight M-F, and they essentially gave him the field. He never declared an emergency.

2. I acknowledge the wiring problem as a factor, but firmly believe this aircraft was in trouble when it touched down due to the tailwind and that decision was the Probable Cause. For the record, I am not the only person who feels this way.

3. According to the NTSB, the pilot in command (PIC) for Flight 3296 had an estimated 20,000 hours total flight time, but only 101 hours in the Saab 2000 aircraft (the co-pilot, with 1,446 hours total time, had 147 hours in the aircraft). In PenAir’s General Operating Manual, and dating back to the company’s previous ownership, PICs were required to have 300 hours minimum in the Saab 2000 before operating into Dutch Harbor. (Similar requirements have existed forever for other companies operating there.) Per the manual, flight time minimums could be waived only if approved by the company Chief Pilot. No such waiver was found during the investigation and one month after the accident, the Chief Pilot was replaced.

4. I do love the FedEx statement on this crash: “FedEx Flight 8312, a Cessna C208 feeder aircraft operated by CSA Air, en route from Milwaukee to Rhinelander, Wisconsin, experienced an incident upon landing. The pilot was taken to a local hospital for observation. We are thankful for the first responders who were dispatched to the scene, and we are fully cooperating with investigating authorities.” The words “incident upon landing” are doing A LOT of work.