Close call at Paris airport

[atrib]

[YOUTUBE]https://www.youtube.com/watch?v=AM01NSZyA7I[/YOUTUBE]

This is the kind of stuff that scares me as a private pilot. It can happen at any airport, but is more likely at bigger airports with multiple runways and multiple towers handling arrivals and departures.

United 57 from Newark was on final approach for Runway 09L (9 LEFT) at Charles de Gaulle Airport (CDG) and had been cleared for this approach by Paris Center (or whatever they call it in Europe). CDG was doing departures on 09R and arrivals on 09L that day. The two runways are parallel and are located in close proximity to one another. United 57 was handed off to CDG Tower on final and the Tower controller cleared him to land 09R (the runway handling departures). This was likely a slip of the tongue, as 09R was not clear at the time. A flight had just landed and was yet to clear the active 09R runway markers. And the Tower had instructed a second aircraft (Easyjet 320) to line up and wait on 09R. The United 57 pilots accepted this change and made a turn to line up with 09R as instructed, and continued to fly the approach.

The Easyjet pilot crossed the 09R hold short line and entered 09R as instructed. As part of routine procedure, the pilots checked the extended centerline on the runway to make sure the runway and approach was clear. The Easyjet pilot saw United 57 on final approach less than a mile from the threshold and issued a clear, unambiguous instruction on the radio for United 57 to cancel the approach and go around. United 57 did a go around (they had also seen the Easyjet enter 09R) , and crossed the 09R threshold (the end of the paved runway) just 80 feet above ground level.

Had this been a foggy day, or if the pilots on the ground had not been paying close attention, this could have turned out quite differently. There are electronic runway incursion systems to act as safeguards (and in this case they sounded the alarm too late), but it is always the responsibility of the humans, the controllers and the pilots to make sure that everything is as should be. A simple mistake from the controller could have resulted in a tragedy but it didn't because the humans were paying attention.

 

[Swammerdami]

I am not a pilot but still find Mentour Pilot's analyses very interesting; I'd already seen this one yesterday.

Mentour assigned zero blame in this incident, excusing the controller because skills get rusty due to lowered activity during the pandemic. (I'll guess Mentour has a high threshold for assigning blame to any fellow professional.) With 2 pilots and 2 controllers acting promptly, Mentour treats the incident as a success story!

The United flight (piloted by Americans) requested confirmation of the unusual side-step: "Understand cleared to land 09 Right, side step for 9 Right United 57"; they received no response to this, but continued with the side-step approach. Mentour claims that two cultural differences may have played a role. (a) Side-steps are much more common at airports in the U.S. than in Europe. (b) The word "Confirm" would have alerted the European controller better than the American "Understand." (Better yet would have been for United 57 to signal just "Confirm clearance" and force the controller to repeat the instruction.)

"Confirm" apparently is a question word, while "Understand" is mostly just a confirmation that the instruction has been heard. (Mentour doesn't play the actual audio; does intonation matter?) As an ignorant layman, I frequently find these audio signals to be ambiguous: are they questions or not? Why not use a less ambiguous phrase, e.g. "Please confirm"?

 

[atrib]

I put the responsibility for this incident on the Tower controller, first for clearing United to land on the wrong runway, and second, for not responding to, and correcting the United pilot's readback of the landing clearance. I think the United pilot did what he was supposed to do - he read back the instructions and even told the Tower that they would be doing a side step to 09R to reflect the change from 09L to 09R. United's readback was not subtle or easily misunderstood, and the Tower should have picked up on the error and corrected it at this point, since United was clearly changing their approach to another runway.

I'm not sure United was asking for confirmation from the Tower. I think they were simply reading back the Tower's instructions. In the US it is not that uncommon for aircraft to get moved to another runway while they are inside the approach markers, and US pilots train for such scenarios. But the way it is done is that the controller will usually ask the pilot if they can accept the change to the new runway, especially if they are already stabilized on the approach. That did not happen here.

The Approach and Tower controllers at big airports are extremely busy, and commercial pilots flying into these airports understand that, and limit their radio talk to the minimum required. Also, United 57 was probably stabilized at the middle marker (or closer) for 09L at the time the Tower issued the landing clearance, and they didn't have a lot of room to change to the new approach. The pilot flying would have been focused on the task of canceling the programmed approach, and hand-flying the aircraft to get established on the new approach.

 

[Swammerdami]

If it's OK with atrib, I'd like to piggy-back on this thread to mention  Air Astana Flight 1388, one of the most amazing air-flight incidents I've ever heard of.

Show

The plane was miswired: left aileron cable connected to where the right aileron cable should go; and vice versa. This might not have been too bad — just turn the yoke left to roll right and vice versa — but the yoke also controls the spoilers which were NOT miswired.

Like United 57, it was covered by Mentour Pilot:
[YOUTUBE]5ywaMkMTwWk[/YOUTUBE]

 

[atrib]

Great job by the flight crew. Must have been exhausting. Is the CVR recording available online? I did a quick search but couldn't find it.

 

[Swammerdami]

This appears to be Part 2 of the recording
[YOUTUBE]evYLkhxoP3U[/YOUTUBE]

 

[Swammerdami]

I'll offer a renewed plug for Mentour Pilot's videos. All of his videos are interesting but I'll link to his discussion of the 1979 tragedy of AA Flight 191. I am impressed with how thoroughly that accident was studied, and how much detail Mentour Pilot gives. (I have my own intricate techie stories to tell about computers, but I'm afraid they'd bore everyone to tears.)

I linked to AA191 because of a personal anecdote. After that crash, DC-10's were barred from flying in U.S. airspace for several weeks; I was on one of the first flights (SF to NY, IIRC) after the DC-10 was re-certified.

While waiting to board — we could see the aircraft through the windows — some of the passengers mentioned that we were boarding a DC-10. One lady refused to believe it: The airline couldn't be putting her on an unsafe plane. "That's a DC-10, lady." She refused to believe them and boarded anyway. I was seated near her; as the DC-10 was taxiing for take-off she consulted the little brochure, finally realized she was on a DC-10 and buzzed the stewardess that she wanted off! The pilot turned around and went back to the gate, announcing on the intercom that "One of the passengers has decided not to fly with us." Everybody groaned. By the time we got back to the gate, the lady was too embarassed to disembark so we taxiied back and took off with her on-board!

ETA: If it's OK with atrib, I'll test my super-powers and change the thread title to be more inclusive.

 

[TV and credit cards]

Interesting that airline maintenance personnel can change the maintenance procedures of the manufacturer. There should be one controlling authority. Airline maintenance crews should submit feedback requesting the change and the manufacturer would put out a revised procedure as necessary.
Excessive maintenance and unnecessary steps in the procedure are a big issue. They invite mistakes and add to cost. I can understand why they would not want to remove so many connectors.

 

[Swammerdami]

I continue to watch Mentour Pilot's YouTubes. I am not a pilot, and am boggled by how complicated it is to fly a large modern airplane, but Mentour Pilot does a great job of explaining incidents in huge detail.

Human performance is often key to an incident. Sometimes heroic pilots salvage an airplane whose hardware has gone very wrong In other incidents, like Airblue Flight 202 — the latest YouTube on the channel — there is nothing wrong with the hardware but human actions lead to tragedy.

It is often the interaction between the TWO pilots that is most important. BOTH pilots will need to work very hard, and cooperate, in complicated situations; Mentour reports many such happy incidents. What does NOT work — and has been the cause of several accidents reported by Mentour Pilot — is when, early in the flight, the senior captain insults the skills of a much younger and less experienced First Officer. That co-pilot becomes intimidated and reluctant to correct the Captain if/when the Captain makes mistakes or goes bonkers. (All or almost all the videos I've noted where the cockpit has this problem are with South Asian or Southeast Asian personnel.) In the case of Airblue 202 the captain was 61 years old, twice the age of the F.O. The captain had recently left his previous airline with its mandatory 60-year retirement age. (I wonder if Airblue tightened its age requirement after this flight.)

Islamabad had bad weather on the day of Airblue 202 and tailwind prevented landing on runway 30. Instead a very difficult landing on runway 12 was required. The captain's performance was abysmal; and the co-pilot was too intimidated to take control.

One of several mistakes made was that the Captain, needing to turn left, changed the auto-pilot heading but this was ignored because he neglected to pull the appropriate mode switch. (It would seem simpler to just turn the control wheel left, but what do I know.) Since the autopilot ignored this heading, the captain just kept decreasing the target heading! When he finally pushed the proper switch the targeted heading was more than 180° less than the current heading, so the auto-pilot did a right turn instead of a left turn! (This was after several "Terrain ahead" warnings — a warning you NEVER want to hear.)

 

[Harry Bosch]

I put the responsibility for this incident on the Tower controller, first for clearing United to land on the wrong runway, and second, for not responding to, and correcting the United pilot's readback of the landing clearance. I think the United pilot did what he was supposed to do - he read back the instructions and even told the Tower that they would be doing a side step to 09R to reflect the change from 09L to 09R. United's readback was not subtle or easily misunderstood, and the Tower should have picked up on the error and corrected it at this point, since United was clearly changing their approach to another runway.

I'm not sure United was asking for confirmation from the Tower. I think they were simply reading back the Tower's instructions. In the US it is not that uncommon for aircraft to get moved to another runway while they are inside the approach markers, and US pilots train for such scenarios. But the way it is done is that the controller will usually ask the pilot if they can accept the change to the new runway, especially if they are already stabilized on the approach. That did not happen here.

The Approach and Tower controllers at big airports are extremely busy, and commercial pilots flying into these airports understand that, and limit their radio talk to the minimum required. Also, United 57 was probably stabilized at the middle marker (or closer) for 09L at the time the Tower issued the landing clearance, and they didn't have a lot of room to change to the new approach. The pilot flying would have been focused on the task of canceling the programmed approach, and hand-flying the aircraft to get established on the new approach.

Yes, you are correct. The local controller at the tower was entirely at fault. I was a tower controller many years ago (worked approach for 6 months also). Good analysis on your part. You're spot on

 

[atrib]

I continue to watch Mentour Pilot's YouTubes. I am not a pilot, and am boggled by how complicated it is to fly a large modern airplane, but Mentour Pilot does a great job of explaining incidents in huge detail.

Modern jets are incredibly complicated, but in reality, flying a modern jet is less complicated than flying a twin-engine Cessna from the 60's or 70's about 98% of the time, that is, until something goes wrong. Pretty much everything on the modern airplane is automated and computer controlled, and everything other than takeoff and the initial climb to the departure fix, and the final 500 feet or so of the descent is flown by the computer flight directors, with occasional inputs from the pilots. The pilots only take over manual control when some significant discrepancy or problem is noted. That is when it gets complicated.

Islamabad had bad weather on the day of Airblue 202 and tailwind prevented landing on runway 30. Instead a very difficult landing on runway 12 was required. The captain's performance was abysmal; and the co-pilot was too intimidated to take control.

One of several mistakes made was that the Captain, needing to turn left, changed the auto-pilot heading but this was ignored because he neglected to pull the appropriate mode switch. (It would seem simpler to just turn the control wheel left, but what do I know.) Since the autopilot ignored this heading, the captain just kept decreasing the target heading! When he finally pushed the proper switch the targeted heading was more than 180° less than the current heading, so the auto-pilot did a right turn instead of a left turn! (This was after several "Terrain ahead" warnings — a warning you NEVER want to hear.)

It is incredibly difficult for me to believe that an experienced pilot with 30,000+ hours would willingly and callously disregard the published approach to a runway while flying in IMC or marginal weather with a low ceiling, in an area with major terrain conflicts within miles of the approach fixes. And he did this while apparently not paying attention to the instruments, and ignoring all the training he must have received to get his commercial license and to get rated on this aircraft. If you get a Terrain warning on your TCAS you switch off AutoPilot, set thrust for TOGA, and pitch for maximum climb. This is a memory checklist every pilot is taught during training. Yet this pilot failed to do even this basic minimum when confronted with this situation. It boggles my mind.

 

[Jimmy Higgins]

I watched several YouTube videos on some of this stuff. It blows my mind how ATC works at all within a normal airport, forget one that is heavily travelled.

 

[atrib]

I watched several YouTube videos on some of this stuff. It blows my mind how ATC works at all within a normal airport, forget one that is heavily travelled.

It is a stressfull job at the bigger airports, even with all the electronic aids available today. Someone made a movie called Pushing Tin which deals with the lives of a group of people working TRACON (Traffic Radar Approach Control) into the NY airspace, and it was very realistic. Not everyone can do that job. Working ATC at smaller airports (like Savannah or Myrtle Beach or even Columbia-CAE) is not that bad because there is very little traffic.

 

[atrib]

Had my first go-around for a runway incursion on Saturday. We were on final to Rwy 20 at JWN in Nashville, cleared to land by the tower. Wife was flying, I was working the radio. A Citation jet that had been holding at 20 suddenly crossed the hold-short line and moved forward into the active runway about 10 to 15 feet. Kate and I called go-around at the same time, she moved the throttles to TO and slowly pulled up the nose. Few seconds later we crossed the threshold at about 200-ft AGL and climbing slowly. I got on the radio and told the tower we were going around as runway was not clear. The tower controller got on the radio and yelled STOP, STOP to the Citation, and later gave him the tower number to call for a possible pilot deviation. We weren't in any real danger of a collision, it was daylight and clear with excellent visibility. We flew the right pattern over the river and came back to land on 20, this time without incident. I called the tower after we were parked at the FBO and told them what we had seen, and they said they would get back to me about the incident.

Overall, I would say this went about as well as one could hope for. Both pilots were paying attention, and Kate did everything right. It was a lot less dramatic than I would have imagined, although she did get a little stressed about it later when we were driving to her mom's place. I was pilot flying on the return leg to Owens and that was totally incident free. The mountains were beautiful as always, and the colors were spectacular, if a little muted compared to last year. I need to motivate myself to buy a couple of go-pros and start filming flights. Maybe start a Youtube channel.

 

[Harry Bosch]

I watched several YouTube videos on some of this stuff. It blows my mind how ATC works at all within a normal airport, forget one that is heavily travelled.

It is a stressfull job at the bigger airports, even with all the electronic aids available today. Someone made a movie called Pushing Tin which deals with the lives of a group of people working TRACON (Traffic Radar Approach Control) into the NY airspace, and it was very realistic. Not everyone can do that job. Working ATC at smaller airports (like Savannah or Myrtle Beach or even Columbia-CAE) is not that bad because there is very little traffic.

Yes. Pushing Tin is the most realistic movie that I've ever seen regarding ATC. Clearly they had controllers advising them on the lingo and how it works. BTW: controllers make more salary when they work at busier areas (probably not surprising).

 

[atrib]

BTW: controllers make more salary when they work at busier areas (probably not surprising).

What would a controller working approach at a busy airspace like Atlanta Center or NY TRACON make today? Do you know? I know someone who works the tower at Columbia (CAE), and he said he makes $120K.

 

[Harry Bosch]

BTW: controllers make more salary when they work at busier areas (probably not surprising).

What would a controller working approach at a busy airspace like Atlanta Center or NY TRACON make today? Do you know? I know someone who works the tower at Owens (CUB), and he said he makes $120K.

Well, I was a controller 30 years ago! I was a tower controller mostly. I did have a final approach rating also. I maybe worked final approach 1 day a week roughly. But radar controllers usually make more than tower as it is a more complex position (at least it was!) and is more difficult. I'd say that tracon in a busy location would make around $160k a year. But that's my guess.

 

[Swammerdami]

I continue to watch Mentour Pilot. The stories are very interesting and the M.P. is very thorough. (I would prefer 15-20 minute show instead of 25 minutes plus ad. M.P. doesn't quite make my Top Four YouTube Narrator's List, but he would with slightly shorter shows.

Can pilots answer my question?
When a plane is configured for auto-pilot, will forceful manual moves — large turn of control wheel, or forceful push or pull on engine thrust levers — have the desired effect; that is, will they over-ride or disable whatever the automatic pilot thinks it's doing? I realize there are a plurality of automatic modes; please make any answer thorough!

The questions arise from two incidents seen on Mentour Pilot. In one, the pilots want to bank to the left sharply, are bewildered when the autopilot doesn't do that in response to a heading keyed into the auto panel, but never actually turn the control wheel counterclockwise. Does that make sense?

In the other incident, pilots are starting Go-Around maneuver which automatically dictates maximum thrust. Yet the landing gear briefly touched the runway (with pilots barely noticing if at all), resetting a latch in the autopilot software, and it does not deliver automatic TOGA thrust. Why didn't the pilots just push the thrust levers to the maximum? Their workload was so high that they didn't notice the levers set for no or low thrust?

The Mentour Pilot I watched most recently was a sad story (from 30+ years ago). The plane with dozens of passengers was making a 40-minute hop from a very minor airport in Brazil to Benin, North by Northeast (bearing 027 degrees). The pilot had called in ill for a recent briefing where it was explained that a new system would provide a decimal point: "0270 degrees" would mean 27.0 degrees. The pilot, having missed the briefing, dialed in 270 for the heading -- due West. (Does it seem odd that the pilot either didn't know the general heading to Benin, or was innumerate enough that it didn't "register" that 270 was due West?)

Protocol required that the co-pilot consult the flight plan independently to set the heading on the co-pilot's panel. But instead he simply copied the pilot's setting. They flew due West, were surprised when Benin didn't appear, made a 180-degree turn still looking for Benin, mistook a North-South River for the East-West Amazon!, and finally figured out what had happened.

Even after realizing their blunder, they missed a nearby airport and flew in the jungle until fuel was exhausted. Most souls survived. Pilot and co-pilot were sentenced to prison terms.

 

[atrib]

I continue to watch Mentour Pilot. The stories are very interesting and the M.P. is very thorough. (I would prefer 15-20 minute show instead of 25 minutes plus ad. M.P. doesn't quite make my Top Four YouTube Narrator's List, but he would with slightly shorter shows.

Can pilots answer my question?
When a plane is configured for auto-pilot, will forceful manual moves — large turn of control wheel, or forceful push or pull on engine thrust levers — have the desired effect; that is, will they over-ride or disable whatever the automatic pilot thinks it's doing? I realize there are a plurality of automatic modes; please make any answer thorough!

Generally, yes - any input on the yoke, trim and thrust/throttle will override the AP, and potentially even switch it off. The AP on modern turbofan and turboprop aircraft is complex, each system is different, and the response of the AP will be a function of how the flight control computers are programmed and the mode it is in. On my Cessna 414a running an aftermarket AP (the Garmin GFC 600 that ties into the Garmin G600 panel), the AP controls everything but the throttle and fuel mix, i.e. the computer can automatically control the pitch, roll, pitch trim, staggered descent with rate, TO/GA, LVL, but the throttle has to be set by the pilot (no hydraulic FADEC). The AP can be set up to either disconnect when it detects a manual input on the yoke and trim, or to go to resume once the manual input has been completed (useful when you need to make small heading deviations to avoid weather and you use the yoke instead of the panel). On modern jets, the AP is also usually programmed to automatically turn off with a warning when certain operating parameters are exceeded, such as when the airplane experiences heavy turbulence or when there is a problem.

 

[Jimmy Higgins]

Darn it! Atrib was our go to for flight too.

Austin Airport was bloody close to a disaster. It sounded like the controller tried to stuff a Southwest takeoff into the schedule, instead of waiting for the Fedex plane to land. Both planes did as ordered and it would have led to a collision had the Fedex plane not noted the Southwest plane, which based on what I've seen about this incident, wasn't something that should have been expected to happen. It likely occurred because Fedex asked for confirmation about clearance, which meant they had doubts about Southwest being out of the way in time. From the numbers, it appeared Fedex was darn close to the ground to. And all of this, in the fog!