Near miss with 5 airliners waiting for T/O on taxiway "C" in SFO!
Join Date: Jan 2011
Location: Brasil
Age: 42
Posts: 145
Likes: 0
Received 0 Likes
on
0 Posts
2 observations from this. I'm not being flippant, I am interested to know the purpose of the "unknown" fields in the document. Can the number of unkown injuries ever exceed zero? What exactly is this field used for?
The part (I paraphrase) "The closest lateral distance bectween AC and any aircraft was 29 feet"... Is that distance between the wingtips or distance from cockpit to cockpit, do we also know whether this distance was between C and 28R or on the other side of C from 28R? looking at the photos and video of a landing on 28R at KSFO, you can see four or five sitinct red lights to the right of the frame, given that the aircraft had already passed "over" taxiway C, was the aircraft to the right and heading to these lights?
Note that aircraft symbols on WebTrak are generic ones and not specific to type or to scale.
Join Date: May 2009
Location: Confoederatio Helvetica
Age: 69
Posts: 2,847
Likes: 0
Received 0 Likes
on
0 Posts
From NavCan.
TSB Report#A17F0159: C-FKCK, an Airbus 320-200 aircraft operated by Air Canada, was conducting flight ACA759 from Toronto, ON (CYYZ) to San Francisco, CA (KSFO). As the aircraft was on a visual approach to Runway 28R at KSFO, ATC cleared ACA759 to land. Approximately 0.6 nautical mile from the runway threshold, the flight crew asked ATC to confirm the landing clearance for Runway 28R because they were seeing lights. ATC responded in the affirmative, and re-cleared ACA759 to land on Runway 28R. The controller was coordinating with another facility when a flight crew member from another airline taxiing on Taxiway C queried ATC as to where ACA759 was going, then stated that ACA759 appeared to be lined up with Taxiway C which parallels Runway 28R. ACA759 had overflown Taxiway C for approximately 0.25 miles when ATC instructed the aircraft to go around. Four aircraft were positioned on Taxiway C at the time of the event. It is estimated that ACA759 overflew the first two aircraft by 100 feet, the third one by 200 feet and the last one by 300 feet. The closest lateral proximity between ACA759 and one of the four aircraft on Taxiway C was 29 feet. The NTSB is investigating.
UPDATE: FAA Report: The following information was reported by FAA Washington Operations Centre: On July 7 at 23:56 PST (July 8 2017 at 02:56 EDT) Air Canada flight 759, an Airbus A320 from Toronto (CYYZ) to San Francisco (KSFO) was cleared to land runway 28R and instead line up for Taxiway C which is parallel to the runway. The aircraft overflew United 1 and Philippine Airlines 115 by 100 feet, United 863 by 200 feet and United 1118 by 300 feet before being issued a go around by Air Traffic Control.
UPDATE: FAA Report: The following information was reported by FAA Washington Operations Centre: On July 7 at 23:56 PST (July 8 2017 at 02:56 EDT) Air Canada flight 759, an Airbus A320 from Toronto (CYYZ) to San Francisco (KSFO) was cleared to land runway 28R and instead line up for Taxiway C which is parallel to the runway. The aircraft overflew United 1 and Philippine Airlines 115 by 100 feet, United 863 by 200 feet and United 1118 by 300 feet before being issued a go around by Air Traffic Control.
Join Date: Dec 2006
Location: Florida and wherever my laptop is
Posts: 1,350
Likes: 0
Received 0 Likes
on
0 Posts
jack11111
It would really do everyone a lot of good to read about visual perception and visual cognition. Our brains invent a huge amount of what we think we are 'seeing' much of the time our eyes are scanning around but the brain uses the informaiton from the balance system to edit out the movements of the eyes (see http://thebrain.mcgill.ca/flash/caps...e_movement.pdf )
(See http://www.drjoebio.com/uploads/1/8/..._our_midst.pdf )
The above paper includes the well known cognition experiment where a basket ball game is played back on video and observers asked to count the number of passes as the game goes on a person wearing a gorilla suit walks through the players across the screen - the gorilla is almost never noticed.
I remember from a long time ago an incident where an aircraft taxiing taxied into a large red British post office truck lost on the taxiway - the pilot 'didn't see it'
There are two effects here - something that is unexpected may be edited out by the cognition process in the brain that includes extra things happening or seeing things happen because they are expected to happen. So I expect to see 2 runways so I see 2 runways, I do not expect to see a large red cross - so I see no large red cross.
Add to this the effect of workload on a challenging approach that tends to lead to 'cognitive tunneling' or fixation on particular tasks and it is unsurprising that what should be obvious is not. All of these effects are exacerbated by fatigue and circadian stress making them far more likely.
It is really worth reading the two references and other similar references to understand what can _and_will_ happen to you.
It is important that ANSPs also realize that reliance on 'large red flashing crosses' will not work. Indeed the more attention getting it is the more likely the brain is to edit it out like the gorilla in a basket ball game. Therefore, the ANSPs that insist that the tower verbally reinforce the warning that the runway configuration is different have got it right - all of a sudden the red-cross is expected and it will be seen.
Don't fly into any gorillas
How can one mistake 28 left is closed or not...there's a big lighted "X" on the threshold.
In these studies, observers engage in a continuous task that requires them to focus on one aspect of a dynamic scene while ignoring others. At some point during the task an unexpected event occurs, but the majority of observers do not report seeing it even though it is clearly visible to observers not engaged in the concurrent task
The above paper includes the well known cognition experiment where a basket ball game is played back on video and observers asked to count the number of passes as the game goes on a person wearing a gorilla suit walks through the players across the screen - the gorilla is almost never noticed.
I remember from a long time ago an incident where an aircraft taxiing taxied into a large red British post office truck lost on the taxiway - the pilot 'didn't see it'
There are two effects here - something that is unexpected may be edited out by the cognition process in the brain that includes extra things happening or seeing things happen because they are expected to happen. So I expect to see 2 runways so I see 2 runways, I do not expect to see a large red cross - so I see no large red cross.
Add to this the effect of workload on a challenging approach that tends to lead to 'cognitive tunneling' or fixation on particular tasks and it is unsurprising that what should be obvious is not. All of these effects are exacerbated by fatigue and circadian stress making them far more likely.
It is really worth reading the two references and other similar references to understand what can _and_will_ happen to you.
It is important that ANSPs also realize that reliance on 'large red flashing crosses' will not work. Indeed the more attention getting it is the more likely the brain is to edit it out like the gorilla in a basket ball game. Therefore, the ANSPs that insist that the tower verbally reinforce the warning that the runway configuration is different have got it right - all of a sudden the red-cross is expected and it will be seen.
Don't fly into any gorillas
The TSB and FAA reports are subtly different.
The TSB reports that the aircraft had overflown Twy C for around a quarter of a mile at the point when ATC instructed a GA.
The FAA state that the A320 had passed over all four queuing aircraft before being told to GA.
That may or may not amount to the same thing, although if FR24 is to be believed there was about 0.4 miles between the first and last queuing aircraft at the point where the ACA flew over (UAL1 was just about to turn the corner towards the threshold and UAL1118 was just past the November link).
That aside, given that both reports confirm the A320 was already climbing by the time it overflew the third aircraft in the queue, it's clear that the GA was instigated by the pilot before the ATC instruction was given. That's consistent with the ATC recording:
SFO Tower: "Air Canada, go around."
ACA759: "In the go around. Air Canada 759."
The TSB reports that the aircraft had overflown Twy C for around a quarter of a mile at the point when ATC instructed a GA.
The FAA state that the A320 had passed over all four queuing aircraft before being told to GA.
That may or may not amount to the same thing, although if FR24 is to be believed there was about 0.4 miles between the first and last queuing aircraft at the point where the ACA flew over (UAL1 was just about to turn the corner towards the threshold and UAL1118 was just past the November link).
That aside, given that both reports confirm the A320 was already climbing by the time it overflew the third aircraft in the queue, it's clear that the GA was instigated by the pilot before the ATC instruction was given. That's consistent with the ATC recording:
SFO Tower: "Air Canada, go around."
ACA759: "In the go around. Air Canada 759."
Ian W,
You speak great sense. Whether or not the illusions you describe are applicable in this incident, it behoves all pilots to study the types of visual illusions that can occur when flying. Take a look at this:-
https://www.youtube.com/watch?v=IGQmdoK_ZfY
You speak great sense. Whether or not the illusions you describe are applicable in this incident, it behoves all pilots to study the types of visual illusions that can occur when flying. Take a look at this:-
https://www.youtube.com/watch?v=IGQmdoK_ZfY
Join Date: Feb 2015
Location: New Hampshire
Posts: 152
Likes: 0
Received 0 Likes
on
0 Posts
Considering these three things:
1) From his remarks, the pilot was observing Charlie closely - seeing what he believed was traffic on that "runway".
2) Images were posted earlier in this thread of roughly what might be seen from the cockpit. Those images should have made the runway situation very clear to the pilot.
3) Before the go around was issued, the pilot had already overflown a quarter mile of runway.
I don't see an easy explanation for this. But here is my best guess:
On first siting of the runways, at long distance, the pilot mis-identified Charlie as 28R. He also saw traffic on that runway. From that point on, he was focused on the status of that traffic. In his mindset, it was more important to resolve the traffic issue that to further peruse the overall runway pattern. (Per a previous post, he wasn't looking for the guerrillas.)
At a certain point on his approach, he decided that the traffic might not resolve in time for his landing - so he informed the tower of the traffic - and was recleared for the landing.
He have had a rapidly improving view of the traffic of Charlie - so it is unlikely (or impossible) at that point that he would have resolved to land. Quite the contrary, he must have been very alarmed at the tower response. It probably indicated to him that they were completely unreliable.
By the time the other craft alerted the tower of the Charlie approach, 759 had no intention of landing and the go around was already in progress. But at that point, he still had a problem: he didn't not know what clearance the other traffic on "28R" (ie, Charlie) had received or would soon receive. So his close approach to the runway was likely intentional - for the purpose of keeping his avoidance options open and to see and be seen.
1) From his remarks, the pilot was observing Charlie closely - seeing what he believed was traffic on that "runway".
2) Images were posted earlier in this thread of roughly what might be seen from the cockpit. Those images should have made the runway situation very clear to the pilot.
3) Before the go around was issued, the pilot had already overflown a quarter mile of runway.
I don't see an easy explanation for this. But here is my best guess:
On first siting of the runways, at long distance, the pilot mis-identified Charlie as 28R. He also saw traffic on that runway. From that point on, he was focused on the status of that traffic. In his mindset, it was more important to resolve the traffic issue that to further peruse the overall runway pattern. (Per a previous post, he wasn't looking for the guerrillas.)
At a certain point on his approach, he decided that the traffic might not resolve in time for his landing - so he informed the tower of the traffic - and was recleared for the landing.
He have had a rapidly improving view of the traffic of Charlie - so it is unlikely (or impossible) at that point that he would have resolved to land. Quite the contrary, he must have been very alarmed at the tower response. It probably indicated to him that they were completely unreliable.
By the time the other craft alerted the tower of the Charlie approach, 759 had no intention of landing and the go around was already in progress. But at that point, he still had a problem: he didn't not know what clearance the other traffic on "28R" (ie, Charlie) had received or would soon receive. So his close approach to the runway was likely intentional - for the purpose of keeping his avoidance options open and to see and be seen.
Guest
Join Date: Apr 2009
Location: On the Beach
Posts: 3,336
Likes: 0
Received 0 Likes
on
0 Posts
Join Date: Feb 2015
Location: New Hampshire
Posts: 152
Likes: 0
Received 0 Likes
on
0 Posts
By my guess: It was intentional in the sense that he had a plan for avoiding traffic - and was following it fairly accurately. And the plan was formulated well before he reached the runway threshold.
However they were all too close for comfort...
So his close approach to the runway was likely intentional - for the purpose of keeping his avoidance options open and to see and be seen.
Join Date: Jun 2001
Location: Rockytop, Tennessee, USA
Posts: 5,898
Likes: 0
Received 1 Like
on
1 Post
Join Date: Jun 2000
Location: last time I looked I was still here.
Posts: 4,507
Likes: 0
Received 0 Likes
on
0 Posts
Before the go around was issued, the pilot had already overflown a quarter mile of runway.
Whether the GA was pilot induced or ATC induced is a debate. The above statement/report causes me a question. (Thanks to those who posted the diagrams. Early in the post there were people talking about 28L, 28C & 28R, I thought. My mistake, as we now see C is to the outside of 28R and not between.)
The threshold 28R and the beginning of taxi C are side by side; or is there a heavily displaced rwy threshold. I've asked what was the PF using for glide path guidance. If it was the ILS 28R then the LLZ deviation should have alerted PF & PM: if it was the PAPI then they are on the left edge of 28R. If AC overflew the taxiway for 400m BEFORE the GA, and at a low height, and if they were on the correct glide path to land the correct distance in, it suggests they flew level for 3-400m. That in itself is an odd manoeuvre. I've missed something earlier. how far down the taxiway were the holding a/c? Were they taking full length or an intersection? If the AC initiated the GA at 100', but after 400m down the taxiway, where was the touchdown aiming point? And at that point the holding a/c must have been very visible.
What happens in USA after an incident like this? Is there an immediate debrief of the crew on site by an FAA or airport official? I'm assuming they night-stopped; is that correct?
Whether the GA was pilot induced or ATC induced is a debate. The above statement/report causes me a question. (Thanks to those who posted the diagrams. Early in the post there were people talking about 28L, 28C & 28R, I thought. My mistake, as we now see C is to the outside of 28R and not between.)
The threshold 28R and the beginning of taxi C are side by side; or is there a heavily displaced rwy threshold. I've asked what was the PF using for glide path guidance. If it was the ILS 28R then the LLZ deviation should have alerted PF & PM: if it was the PAPI then they are on the left edge of 28R. If AC overflew the taxiway for 400m BEFORE the GA, and at a low height, and if they were on the correct glide path to land the correct distance in, it suggests they flew level for 3-400m. That in itself is an odd manoeuvre. I've missed something earlier. how far down the taxiway were the holding a/c? Were they taking full length or an intersection? If the AC initiated the GA at 100', but after 400m down the taxiway, where was the touchdown aiming point? And at that point the holding a/c must have been very visible.
What happens in USA after an incident like this? Is there an immediate debrief of the crew on site by an FAA or airport official? I'm assuming they night-stopped; is that correct?
Guest
Join Date: Apr 2009
Location: On the Beach
Posts: 3,336
Likes: 0
Received 0 Likes
on
0 Posts
Do we know whether the NTSB is even looking at this SFO incident?
No, it isn't. We know (from the TSB report) that the aircraft was already climbing when the controller instructed a GA.
No significantly displaced threshold according to Google Earth or the airport chart.
See above. We don't know at what point the pilot initiated the GA, and we're unlikely to until we see the final report. But we DO know from the TSB that the aircraft had established a positive ROC by the time it overflew UAL863 (the third aircraft in the queue) and maybe even before that. It looks like the heights being quoted have a ±50' resolution, so it may even have been climbing by the time it passed over the second or even the first aircraft (which, scarily, would suggest that it may have been even lower than 100' at that point).
See my previous post. Subject to the usual FR24 caveats, all 4 aircraft in the queue had passed the last link (November) that would have permitted an intersection takeoff.
The threshold 28R and the beginning of taxi C are side by side; or is there a heavily displaced rwy threshold.
If AC overflew the taxiway for 400m BEFORE the GA, and at a low height, and if they were on the correct glide path to land the correct distance in, it suggests they flew level for 3-400m.
Were they taking full length or an intersection?
Join Date: Dec 1998
Location: UK
Posts: 294
Likes: 0
Received 0 Likes
on
0 Posts
Clearly a serious incident but clearly not the first crew to have done it. Quite a few US airports have multiple parallel runways and wide taxiways that can be confused (Las Vegas springs to mind). Confirmation bias of seeing only 1 lit runway to the left. Not all runways have approach lights and centreline lights so a taxiway can look a bit like a runway. The ILS probably wouldn't be tuned as think it's an RNAV type approach.
In some ways I find it harder in daylight into a setting sun to distinguish runways and taxiways, especially as the lights are hard to see then. San Diego has a centre stripe down the middle of the runway which is a different texture which makes it look like a taxiway until you see it has another section either side.
The newest Boeings have RAAS which tells you what you're lined up with on final approach and would warn you if lined up with a taxiway, so technology is catching up.
In some ways I find it harder in daylight into a setting sun to distinguish runways and taxiways, especially as the lights are hard to see then. San Diego has a centre stripe down the middle of the runway which is a different texture which makes it look like a taxiway until you see it has another section either side.
The newest Boeings have RAAS which tells you what you're lined up with on final approach and would warn you if lined up with a taxiway, so technology is catching up.