Cathay messy in SFO

[Officer Kite]

https://www.youtube.com/watch?v=78V66AbPGj4

Well this is worrying, you can almost feel the uncertainty/confusion in the Cathay cockpit, another reminder of how important situational awareness is and to never lose sight of it. Surely this would've been a 3 man ops with someone observing during this phase to keep overall awareness?

With all that high tech in the 350 too

[VH DSJ]

Looks like Cathay went right through the 28L localizer and was intercepting the the 28R localizer. Did they have the correct runway in the box I wonder?

[Locked door]

How can it be that the UAL failed to follow a TCAS RA. Or have I misunderstood?

TCAS is the final barrier to a mid air collision. Not following one is insane.

[Officer Kite]

How can it be that the UAL failed to follow a TCAS RA. Or have I misunderstood?

TCAS is the final barrier to a mid air collision. Not following one is insane.

Strictly speaking that's true and I was always taught the same and wondered too. It seems they took reassurance from SFO tower telling them the traffic was not on a collision course. One does wonder how far Cathay were willing to go on this rollercoaster of an approach before they themselves thought it was time for a GA!

[cabbages]

just looking for that youtube clip. Did this happen yesterday Officer K??

[Meester proach]

Should have just followed the RA , not tried to reason the position of the Cathay

[3Greens]

just looking for that youtube clip. Did this happen yesterday Officer K??
rather worrying that UAL seemingly ignoranced a TCAS RA. Cathay, whilst far from blameless, at least followed his TCAS.
no matter what Atc say, always follow the TCAS. NOT atc.

[gearlever]

just looking for that youtube clip. Did this happen yesterday Officer K??

Video is of 28-08-2019, not stating when it happened.

[tigerinthenight]

The only thing ATC should be saying in the event of a TCAS RA is "roger". Giving traffic information is a big no.

[Fortissimo]

In many parts of the USA, pilots will often operate with TCAS in TA mode to avoid 'nuisance' RA because of parallel runway ops, mixed GA traffic etc. and the operators think this is quite acceptable as the airspace is designated as congested. The chances are that the UAL aircraft never got an RA because the automatics were not enabled. The danger comes when these guys operate over in Europe, which the US regards as congested in all areas, so it is quite possible that one half of a conflict is only receiving TA. So long as the other party follows the RA adequate separation should result, though the experts might be able to point out circumstances where this may not be good enough.

[Airbubba]

The chances are that the UAL aircraft never got an RA because the automatics were not enabled.

On the tape at 4:34 United 1515 reports that he got a climb RA.

[DaveReidUK]

just looking for that youtube clip. Did this happen yesterday Officer K??

It happened at least 3 months ago, judging by the United's flight number.

[moosepileit]

No followee, no guarentee.

[Uplinker]

Sounds like a very bad day in the office for the Cathay.

Long flight, poor crew rest, poor arrival briefing?

[Tomaski]

Can’t speak to the specifics of this incident, but we have SFO specific procedures that allows the Captain to select TA only for TCAS during closely spaced parallel approaches in VMC with discretion to disregard warnings if the encroaching aircraft is in sight and safe separation can be maintained.

[VThokie2]

The only thing ATC should be saying in the event of a TCAS RA is "roger". Giving traffic information is a big no.

I take what you are saying however when faced with a question, ATC tried to help. Instead of:

UAL 1515: “OK,Sir, Where is that aircraft behind us? I got an RA here to climb”

Tower: “Roger”

(Btw “Roger” is the response to “TCAS RA” which UAL did not technically say”)

To add to what someone said earlier .... I’ve flown domestically in the US and as an expat flying heavies, for both operators there was a note in the company pages recommending TA only for KSFO when conducting the 28L/28R approaches. Who knows what UAL recommends in theirs.

The bigger industry wide issue is operators having difficulties flying a visual (based on proceeding traffic, or just the runway in site). I completely get why, since when I was overseas we practiced patterns all the time in the sim so that we could say we could do it, but in the real operational environment no one would dare fly a visual from a downwind, even CAVOK daylight. It’s aviation, crap happens and if all envolved in this reflected on what they could’ve done better then that’s a good outcome. However in some companies and aviation cultures punitive actions would be the result. Which leads to people not willing to accept a visual for fear of cocking it up! Yet puff up their chests and think themselves superior aviators.

[oceancrosser]

Can’t speak to the specifics of this incident, but we have SFO specific procedures that allows the Captain to select TA only for TCAS during closely spaced parallel approaches in VMC with discretion to disregard warnings if the encroaching aircraft is in sight and safe separation can be maintained.

We have identical procedures. The Cathay guys seemed lost in SFO.

[sonicbum]

According to the audio the UAL communicates "I got a RA here to climb" so it looks like they did get a TCAS RA.

[RoyHudd]

RA =OBEY

No issue

[ScepticalOptomist]

I was a little shocked when the UAL says “he’s passing underneath us” and then doesn’t respond to a climb RA.

Without being on the flight deck it’s hard to know exactly what the dynamics looked like but it sounds pretty bad.

Glad it all turned out ok.

[West Coast]

In many parts of the USA, pilots will often operate with TCAS in TA mode to avoid 'nuisance' RA because of parallel runway ops, mixed GA traffic etc. and the operators think this is quite acceptable as the airspace is designated as congested. The chances are that the UAL aircraft never got an RA because the automatics were not enabled. The danger comes when these guys operate over in Europe, which the US regards as congested in all areas, so it is quite possible that one half of a conflict is only receiving TA. So long as the other party follows the RA adequate separation should result, though the experts might be able to point out circumstances where this may not be good enough.



Why the rush to conclusion? Hopefully you saw the numerous references that debunked your "chances are" assumption.

Also, what parts of the USA do crews operate in TA mode? A mate telling you doesn't count.

[cessnaxpilot]

In many parts of the USA, pilots will often operate with TCAS in TA mode to avoid 'nuisance' RA because of parallel runway ops, mixed GA traffic etc. and the operators think this is quite acceptable as the airspace is designated as congested. The chances are that the UAL aircraft never got an RA because the automatics were not enabled. The danger comes when these guys operate over in Europe, which the US regards as congested in all areas, so it is quite possible that one half of a conflict is only receiving TA. So long as the other party follows the RA adequate separation should result, though the experts might be able to point out circumstances where this may not be good enough.


wrong on many points. “The danger comes when in Europe?” Why? ”These guys...”. Seriously? We don’t operate in TA mode like you suggest. There are some GA airports with very closely spaced runways and if you are in day VMC... And you have the traffic in sight... then... you can select TA. This is not as common as you suggest. Or without restrictions, as you suggest.

[hans brinker]

In many parts of the USA, pilots will often operate with TCAS in TA mode to avoid 'nuisance' RA because of parallel runway ops, mixed GA traffic etc. and the operators think this is quite acceptable as the airspace is designated as congested. The chances are that the UAL aircraft never got an RA because the automatics were not enabled. The danger comes when these guys operate over in Europe, which the US regards as congested in all areas, so it is quite possible that one half of a conflict is only receiving TA. So long as the other party follows the RA adequate separation should result, though the experts might be able to point out circumstances where this may not be good enough.

yeah, pull the other one.....
Every 121 operator is required to have TCAS, and not in TA only mode. Every RA is required to be followed (if the UAL crew didn’t they will absolutely get a call from their safety department). When we go to Europe, we behave just like the rest of you.

[ATC Watcher]

As a European-trained Controller I watch that video and R/T with some bemused feelings. Instructing one a/c "maintain visual " and let 2 heavies get so close on radar without saying something is remarkable.
In this scenario only the Cathay could maintain visual since he was behind the United all the time. I guess this situation is a daily thing for the SFO controllers and working well 99% of the time , and basically this is what you get when you want to move the traffic they do in parallel runways.
I flew a couple of times into Oshkosh ( again a few weeks ago) and the approach there not what you will call "European" but it works., and works very well even . Respect..But to get there you need to read a 30 pages instruction booklet beforehand.
I am not so sure the foreign airlines crews operating in say SFO , have all read and understood the booklet ( if there is one) on the local procedures and the differences between US ATC and their home turf.

[Mad As A Mad Thing]

Instructing the aircraft below (CPA) to break off the approach and climb through the aircraft above (UAL) in such close proximity with the higher aircraft completely unable to see the other one seems like a questionable decision by ATC.

[ATC Watcher]

Instructing the aircraft below (CPA) to break off the approach and climb through the aircraft above (UAL) in such close proximity with the higher aircraft completely unable to see the other one seems like a questionable decision by ATC.
I thought so as well (still European thinking) but remember Cathay was told at least twice to maintain visual separation , was definitively behind ,1500 ft lower. 10 kts slower and given a diverging heading , no real chances of collision .

[PerPurumTonantes]

ATC workload? Can't believe someone was actively watching that radar and failing to advise. Is VASAviation simulation accurate?

Perhaps the parallel runways at SFO so close that ATC are used to seeing traces almost on top of each other, hence lack of concern?

Is it just me or have there been rather a lot of incidents recently at SFO? Seem to be a number of factors making it one of the higher risk airports at the moment. What could be done to fix it?

[cappt]

It's not uncommon to get a RA on final in SFO when doing closely spaced visual approaches. Standard procedures are continue if the parallel traffic is in site.
In this case it would have been wise for ATC to break off United and climb them away as it became apparent CPA had not complied with the last instruction and had lost SA.

[Miles Magister]

Guys,
Without reference to the above incident may I suggest that those of you that are not aware of the Eurocontrol ACAS Bulletins take a look at them here https://www.eurocontrol.int/search?keywords=acas+ii+bulletin&sort_by=field_publication_date
They are very good reading and start with No 1. They are published periodically so there are quite a few now and you will not read them all in one sitting but they are really worth reading for all pilots on both sides of the pond.
MM

[George Glass]

cappt, I hope you don’t operate into Sydney. An RA is an RA.
Never ever been trained to second guess it.
What if there is traffic you’re not aware of?

[cappt]

cappt, I hope you don’t operate into Sydney. An RA is an RA.
Never ever been trained to second guess it.
What if there is traffic you’re not aware of?

If there's traffic you not aware of by all means comply with the RA.
SFO 28L/R is separated by 750' between centerlines (enviros won't let them move the runway over). When doing visuals expect to be paired up with parellel traffic. If one aircraft is slightly faster and closing or correcting to their G/S it can trigger a RA. If your told to follow but do not pass that's because your paired aircraft is slightly ahead of you and doesn't have you insight.

[Maninthebar]

If there's traffic you not aware of by all means comply with the RA.
SFO 28L/R is separated by 750' between centerlines (enviros won't let them move the runway over). When doing visuals expect to be paired up with parellel traffic. If one aircraft is slightly faster and closing or correcting to their G/S it can trigger a RA. If your told to follow but do not pass that's because your paired aircraft is slightly ahead of you and doesn't have you insight.


Errrrr how will you are not aware of it if you are not aware of it?

[lambourne]

cappt, I hope you don’t operate into Sydney. An RA is an RA.
Never ever been trained to second guess it.
What if there is traffic you’re not aware of?

In SFO the runways are closely spaced. An RA will
occur due to the simultaneous closely spaced arrivals. ATC gives the instruction “report the traffic in sight”
when the pilot says “in sight” the controller will then state “maintain VISUAL separation cleared for the visual”. At that point the separation is among the pilots. Technique is to turn to TA only to avoid nuisance RA. That is approved UA procedure. If you don’t like that procedure you will go around and around and around on each approach to SFO. Pairing up and flying near the other plane is SOP for SFO. If you can’t fly the plane and adhere to instructions maybe it’s not the airport or the other airlines fault you are having difficulties.

[fleigle]

cappt
Blaming the "enviros" for the runway situation at SFO is totally off course, where would another runway go?.
There is no room at SFO.!
f

[CanadianAirbusPilot]

My airline suggests we use TA only at the several airports like SFO when you get close to other airplanes. If you forgot and you get an RA you are required to follow it. Surprised the UAL guys kept going with CPA flying under them.

Scary hearing the lack of SA onboard CPA. Seems those guys need a bit more time bombing around in a small airplane. When ATC clears them for a visual 28L follow the UAL jet, things should have been easier, but instead it seems it made it harder. Easiest thing should have been, AP off, FD off, A/THR off and fly the airplane. Instead it was painful listening to those guys, sounded very overwhelmed. How much experience are Cathay pilots being hired with these days?

[cappt]

cappt
Blaming the "enviros" for the runway situation at SFO is totally off course, where would another runway go?.
There is no room at SFO.!
f

B/S! Move 28R a couple hundred feet out further into the bay. The current 28R can become bay again with little overall loss of San Francisco Bay.
The elephant in the room here is why can't major international crews fly a visual?

[cessnaxpilot]

I’ve never been anywhere in the world as busy as some US corridors. It’s amazing to me how great the controllers are in New York. You have Newark, Kennedy, La Guardia, and teterboro all within a stones throw. I think there are some regional ATC differences for the environment... but it’s very spelled out on visual approaches, spacing, and TAs. That’s not the difference. SFO is also a busy airspace with OAK and SJC and a lot of GA. CX accepted the visual and they probably shouldn’t have. Tired. Long Duty... sun in the eyes. I’m not sure what happened. ATC and UA crews also interacted. It’s a lot of moving parts and not black and white. What’s missing is the visual from the cockpit. We don’t know what the crews saw or didn’t see. So all the speculation is worthless.

[ironbutt57]

My airline suggests we use TA only at the several airports like SFO when you get close to other airplanes. If you forgot and you get an RA you are required to follow it. Surprised the UAL guys kept going with CPA flying under them.

Scary hearing the lack of SA onboard CPA. Seems those guys need a bit more time bombing around in a small airplane. When ATC clears them for a visual 28L follow the UAL jet, things should have been easier, but instead it seems it made it harder. Easiest thing should have been, AP off, FD off, A/THR off and fly the airplane. Instead it was painful listening to those guys, sounded very overwhelmed. How much experience are Cathay pilots being hired with these days?

personally I suspect the only experience the CX pilots may have been lacking is the experience of sleep

[OldnGrounded]

B/S! Move 28R a couple hundred feet out further into the bay. The current 28R can become bay again with little overall loss of San Francisco Bay. That's just not going to happen. Various versions of the battle have been going on since I was a kid in Foster City, with aircraft on final to 28 L/R skimming overhead "almost within reach." That was a very long time ago.
Even without the objections of the "enviros," the neighbors can be counted on to almost universally, very loudly, and very effectively oppose runway expansion. They always do and, so far, they always win.

The environmental issues are hardly trivial, either. Fully a third of SF Bay has been filled (and much of it been built on) since the mid-nineteenth century. That's a big deal and current law prevents filling more of it without major environmental impact review and restoring previously-filled wetlands elsewhere in the Bay. Difficult and expensive.

There should be plenty of slots available at KOAK and KSJC.

[fleigle]

cappt
"B/S! Move 28R a couple hundred feet out further into the bay. The current 28R can become bay again with little overall loss of San Francisco Bay."
You saying move it north 200'?, where would the GA terminal go?, the Coast Guard Station?, the tank farm to the NW?, the takeoff path would then be even closer to the San Bruno "mountain"...... you are the one with B/S.
f

[pudoc]

Interesting to see not many people blaming the approach controller. How you can plonk an aircraft right below another and ask them to maintain visual separation, whilst at a similar speed and track...legal or not. I know that’s the done thing in the US but there’s a lack of airmanship shown by the controller. Awful vectoring too.

Not a great deal of airmanship shown by either of the flight crews either.

[cappt]

cappt
"B/S! Move 28R a couple hundred feet out further into the bay. The current 28R can become bay again with little overall loss of San Francisco Bay."
You saying move it north 200'?, where would the GA terminal go?, the Coast Guard Station?, the tank farm to the NW?, the takeoff path would then be even closer to the San Bruno "mountain"...... you are the one with B/S.
f

Don’t be silly, the city has been trying to fix the layout since the Willie Brown nineties. The simplest and least invasive solution is move 28R north to the end of the 1’s. The only reason airliners have to do this approach wingtip to wingtip is because of this outdated joke of an international airport. Oh there’s a cloud below 2100’? Your flights now three hour delayed because simultaneous approaches are no longer authorized.

[Backupnav]

Interesting to see not many people blaming the approach controller. How you can plonk an aircraft right below another and ask them to maintain visual separation, whilst at a similar speed and track...legal or not. I know that’s the done thing in the US but there’s a lack of airmanship shown by the controller. Awful vectoring too.

Spot on, pudoc. I think SFO controllers tend to forget that many pilots operate there only now and then, and are far from familiar with the less than optimum airport configuration. Also, expecting a heavy to do a S turn on finals at the end of a long haul flight is unrealistic.
​

[cessnaxpilot]

Interesting to see not many people blaming the approach controller. How you can plonk an aircraft right below another and ask them to maintain visual separation, whilst at a similar speed and track...legal or not. I know that’s the done thing in the US but there’s a lack of airmanship shown by the controller. Awful vectoring too.

Not a great deal of airmanship shown by either of the flight crews either.
from the AIM
A pilot sees the other aircraft involved and upon instructions from the controller provides separation by maneuvering the aircraft to avoid it. When pilots accept responsibility to maintain visual separation, they must maintain constant visual surveillance and not pass the other aircraft until it is no longer a factor. When the CX accepted the visual he was instructed to maintain visual separation but not pass the aircraft. Many airports have very closely spaced runways and the aircraft can fly almost in formation. For awhile they were trying to establish PRM approaches for such airports with a “zone of no transgression” and a monitor approach controller. It’s all in an effort to increase arrival rates.

[pattern_is_full]

C'mon, folks. In 3 pages no-one has mentioned directly that Cathay:

- got and acknowledged an instruction to turn left to 010 - and 38 seconds later was still goofing along on course ~090.
- when reminded of the turn, had forgotten the heading assigned
- was confused about whether he was flying an ILS or visual

I mean, seriously - all you have to talk about is "ATC this" and "United that" and "RA something else?" I don't want to be especially hard on Cathay - but let's not lose sight of the real story here.

[ATC Watcher]

let's not lose sight of the real story here.
Absolutely . The controller just follows the local rules he has been trained for and instructed to do . SFO ,like many other US airports are operated with a waver from the FAA to go below normal separation minima. on APP.
The problem here is that the CX crew appears not prepared ( or even trained) for the unusual approach situation handed over to them , acerbated by the heading problems, those could be explained by fatigue or stress ( look at the R/T speed delivery ) or more likely both combined.
To their discharge, to maintain in a wide-body visual separation with an aircraft less than a Mile away is not something you do everyday outside of the USA..

Now that said, both aircraft always had vertical separation at all times anyway, so nobody was really at risk here , so let's keep things into perspective.

[cactusbusdrvr]

SFO works very well when everyone plays the game right. That means flying the course you are cleared for and maintaining separation that you are responsible for once you are “cleared for the visual.” Or cleared for the FMS visual.

The most formation flying I’ve done is been flying the visual to the 28s in SFO with a 777 on my left side and me keeping station slightly behind to the right. Auto pilot off, auto throttles off. Match speed. Basic airmanship.


ATC relies on the pilots to do as they are cleared. Cathay screwed up several ways. They delayed their base turn, they overshot their course, they failed to slow when they accepted the clearance to follow UAL. ATC did a good job considering that the end result was them breaking off Cathay and sending them around.

United can be forgiven for not responding to the RA immediately for several reasons. They were on short final, inside the bridge and about 3 out. Landing is not when you expect to do a RA maneuver. They heard Cathay acknowledged the speed and maintain visual clearance. They assumed he was complying. They asked tower for advisories on the traffic, they got them.

[Bob Viking]

I find it interesting that many posters have mentioned that the Cathay crew would have been fatigued after a long flight. I have seen similar reasoning on other threads where crews have made mistakes.

Assuming they were within their crew duty times can we really excuse the mistake by saying they were tired?

I know what it’s like to be fatigued but surely the regulations and our best practice should prevent it from causing problems such as this.

If crews are routinely making mistakes due to fatigue then maybe the rules need to be changed.

I realise I am putting the cat amongst the pigeons here but I do love a good debate.

BV

[ACMS]

SFO does seem to have its fair share of issues......

Time to ditch visual approaches for long haul carriers unfamiliar with SFO’s idiosyncrasies.

All fine for local Jet operators not fatigued and close to their body clock....

[Rostermouse]

United can be forgiven for not responding to the RA immediately for several reasons. They were on short final, inside the bridge and about 3 out. Landing is not when you expect to do a RA maneuver. They heard Cathay acknowledged the speed and maintain visual clearance. They assumed he was complying. They asked tower for advisories on the traffic, they got them.

I disagree with absolving UAL here. Perhaps their decision can be understood in the circumstances you describe but their actions were still incorrect. Assuming that other aircraft are follinwg instrucitons correctly, and that ATC have the correct picture, is exactly WHY TCAS was developed...because historically those assumptions have at times proven fatally incorrect.

Doesn't matter where you are on final, doesn't matter if you think you know where the other traffic is or what they are doing - you must comply with a TCAS RA.

The only missing piece I can see here is the the controller didn't instruct a breakout turn for UAL ( I assume SFO were conducting PRM or SOIA?). Even if they had, however, TCAS compliance in the vertical is still expected.

[icarus sun]

If crews cannot fly a heading,and do a visual approach they should not be in any commercial aircraft. Let alone a wide body..

[alf5071h]

#48 “… good debate” :ok:
… often overlooked, hampered by fixed views, and the increasing complexity of aviation operations.
The ‘system’ more often assumes that operations are only ‘complicated’, will always be understood, follow the rules, but in reality the operational environment is ‘complex’, interacting in unforeseen ways and thus unpredictable.

https://blog.usejournal.com/7-differences-between-complex-and-complicated-fa44e0844606

“It is also important to note that in most human systems, complex and complicated co-exist …”

Were both aircraft still on IFR flight plans?
Can ATC arbitrarily change IFR by stating maintain visual?
Would SFO be capable of operating at a high landing rate if IFR was imposed?

edit: #50 “you must comply with a TCAS RA” :ok:

[172_driver]

Were both aircraft still on IFR flight plans?
Can ATC arbitrarily change IFR by stating maintain visual?
Would SFO be capable of operating at a high landing rate if IFR was imposed?

From the AIM in post #44, visual separation may be provided IFR to IFR if the crew report traffic in sight. I am certain none was ever asked to cancel IFR.

[calypso]

I agree a TCAS RA MUST be followed. For start it might not even be based on to the conflicting traffic that you are thinking about. It might be generated by a third traffic you are not aware of.

[sonicbum]

Flying to busy US airports for non US based carriers requires a thorough understanding of the FAA differences with one's own national regulations. As an example for a visual approach according to the FAA :

Controllers may initiate, or pilots may request, a visual approach even when an aircraft is being vectored for an instrument approach and the pilot subsequently reports:


The airport or the runway in sight at airports with operating control towers.
The airport in sight at airports without a control tower.

in EASA land (for example) :

The initial and intermediate approach phases of an approach executed under the direction of a controller comprise those parts of the approach from the time vectoring is initiated for the purpose of positioning the aircraft for a final approach until the aircraft is on final approach and:
(a) established on the final approach path of a pilot-interpreted aid; or
(b) reports that it is able to complete a visual approach; or
(c) ready to commence a surveillance radar approach.


Basically the big difference is that under EASA the visual approach will always come upon a request from the pilot whether in the States You can get one (and You will) by the controller's initiative. Being ready for it is what makes the difference.

[FLCH]

Spot on, pudoc. I think SFO controllers tend to forget that many pilots operate there only now and then, and are far from familiar with the less than optimum airport configuration. Also, expecting a heavy to do a S turn on finals at the end of a long haul flight is unrealistic.
​

Unrealistic ? Unrealistic to have the potential for a go around either ? If you can't do a simple S turn on final long haul or not, you shouldn't be near the controls of a jet.
You get paid to fly an aircraft all you have to do is complete that task.

[alf5071h]

sonic, #55,
In FAA land, irrespective of who initiates a visual approach - need to see the airport, etc, there is no statement of change of responsibility for aircraft separation when flying an IFR flight plan.
Or does the FAA ‘visual approach’ imply visual separation; if so then all relevant aircraft positions should be notified, and those involved must be able to see each other and should manoeuvre to ensure no ACAS resolution conflict.
Gross assumptions all round.

[pudoc]

from the AIM A pilot sees the other aircraft involved and upon instructions from the controller provides separation by maneuvering the aircraft to avoid it. When pilots accept responsibility to maintain visual separation, they must maintain constant visual surveillance and not pass the other aircraft until it is no longer a factor.
When the CX accepted the visual he was instructed to maintain visual separation but not pass the aircraft. Many airports have very closely spaced runways and the aircraft can fly almost in formation. For awhile they were trying to establish PRM approaches for such airports with a “zone of no transgression” and a monitor approach controller. It’s all in an effort to increase arrival rates.
This is why I said "legal or not" in my post. There's no point quoting the AIM, there's a distinct lack of airmanship by all parties. The book of rules won't save you from a disaster, airmanship will. Just ask Sully.

[sonicbum]

Unrealistic ? Unrealistic to have the potential for a go around either ? If you can't do a simple S turn on final long haul or not, you shouldn't be near the controls of a jet.
You get paid to fly an aircraft all you have to do is complete that task.

You get paid to fly an aircraft according to Your operator specific requirements and You are entitled by all means not to perform any manoeuvre that You are not comfortable with, or haven't practiced including S turns on final with a 300T passenger jet. In this event everybody has their share of responsibility but again when operating to busy US airports You need to be ready for visuals, (very) late runway changes with VOR approaches and so on, that's how it works on the other side of the pond, fair enough.

[sonicbum]

sonic, #55,
In FAA land, irrespective of who initiates a visual approach - need to see the airport, etc, there is no statement of change of responsibility for aircraft separation when flying an IFR flight plan.
Or does the FAA ‘visual approach’ imply visual separation; if so then all relevant aircraft positions should be notified, and those involved must be able to see each other and should manoeuvre to ensure no ACAS resolution conflict.
Gross assumptions all round.

I agree with You, but that's why You need to be briefed on how things work at certain airports to avoid any type of issue and trying to comply with ATC requests as far as practical. It's how it works there, it's either You get ready for it or get caught by it.

[RetiredBA/BY]

I find it interesting that many posters have mentioned that the Cathay crew would have been fatigued after a long flight. I have seen similar reasoning on other threads where crews have made mistakes.

Assuming they were within their crew duty times can we really excuse the mistake by saying they were tired?

I know what it’s like to be fatigued but surely the regulations and our best practice should prevent it from causing problems such as this.

If crews are routinely making mistakes due to fatigue then maybe the rules need to be changed.

I realise I am putting the cat amongst the pigeons here but I do love a good debate.i(

BV
You may have been fatigued BUT do you have any experience of long, often night, flights with big time changes and with your circadian rhythm shot to pieces. Max FDP .minimum rest, do it all again as the companies are all working you to the limits.

My worst, LHR to NBO, via CDG, or ADD, dep 2200, 12 hours rest then back to London, all legal, Arr LHR feeling rather dismal to say the least. I declined my company request to op as 767 Captain, AUH to PER ! And we had three crew in those days !

So, many pilots now transferring to part time, no surprise there.

If you have not experienced it perhaps, best to stick to military, Jag, Hawk, matters !

[alf5071h]

sonic #60, thank you; we agree about the issue, but not necessarily the mechanism of how operations work or should work.
An international industry cannot depend on local knowledge, or rely on extensive briefing sheets specific to certain airports, and particularly on obscure, unpublished assumptions or interpretations of general regulations.

Safety is defined by what is done, what is knowable; we have safety margins and adaptability for surprise, but these must not be taken for granted as a normal operation.
If we do, then when there is a challenging situation, as in this incident, the affect of degraded safety margins cross recognised and acceptable safety boundaries.
These ‘normal’ operations are unsafe.

[Bob Viking]

That’s not debate. That’s just telling me to mind my own business.

Are you trying to tell me that at the end of long sectors pilots are always tired and, as a result, are not able to perform a safe approach?

If this is the case then why do the rules allow it?

I suppose the greatest irony here is that, in a Hawk, I completed an approach to 28R at SFO at the end of a 12 hour work day in 2012. So you see, it wasn’t in a wide body, but I have actually been there. I was tired too.

I realise you have a lot of experience in the commercial world (whereas all of my experience is fast jet) and I don’t but I do fly as a long haul passenger quite regularly. If the ability to fly a safe approach is truly so compromised by fatigue then surely something needs to be done about it.

BV

[Uplinker]

.................Are you trying to tell me that at the end of long sectors pilots are always tired and, as a result, are not able to perform a safe approach?

If this is the case then why do the rules allow it?

I suppose the greatest irony here is that, in a Hawk, I completed an approach to 28R at SFO at the end of a 12 hour work day in 2012. So you see, it wasn’t in a wide body, but I have actually been there. I was tired too.

I realise you have a lot of experience in the commercial world (whereas all of my experience is fast jet) and I don’t but I do fly as a long haul passenger quite regularly. If the ability to fly a safe approach is truly so compromised by fatigue then surely something needs to be done about it.

BV





Let’s calm down a second. I for one basically agree with you.

Not always, but yes, myself and others are saying that fatigue sometimes plays a part. We were all unhappy when EASA rules came in, but sadly not enough of us nor BALPA nor our CAA were prepared to refuse the new FTL rules. So here we are. I put in fatigue reports as many others do, but companies are continually screwing more and more hours out of fewer and fewer pilots because the passenger will usually only pay for the cheapest fight. In my company, we fly three crew on flight times over 10 hours or so, so at least we each get a couple of hours’ rest outside the cockpit. However, for us that rest is in a seat in the passenger compartment, with a curtain around it. You can still hear - and are kept awake by - passengers around you talking.

We have video briefings about certain airports but sometimes it can be your first time there, and no amount of words on a page or videos can fully prepare you for an American airport..............

This Cathay crew screwed up. They missed a turn. I don’t know why, but very probably it was fatigue or inadequate preparation or briefing. The Cathay pilot on the radio sounds stressed to me. Maybe none of the flight crew had been to SFO before.

Some Long-haul pilots are short on manual handling skills. They shouldn’t be, but many are. If you only get two landings a month and are encouraged by your airline and FDM to fly with the automatics then this can be the result. I have suggested several times on other threads how this might be addressed.

Bear in mind that airliners are mostly flown by average pilots - (and I include myself in that analysis). Fast jet pilots from the forces are much better pilots than most of us because they were the best 0.0001% (or whatever) selected from all applicants.

UK ATC is the best in the world in my humble opinion and 18 years’ commercial flying. Many US airports would benefit by learning from UK trained and experienced ATCers such as those from Heathrow, Gatwick or Swanwick.

Fly safe :ok:

[OldnGrounded]

As I read this thread, I was reminded of a study from a decade or so ago, entitled "Trajectory Clustering and an Application to Airspace Monitoring (https://www.researchgate.net/publication/45897875_Trajectory_Clustering_and_an_Application_to_Airspac e_Monitoring)." Actually, what I remembered is a a graphic, based on three months of Northern California TRACON records for the Bay Area. Here it is:


https://cimg3.ibsrv.net/gimg/pprune.org-vbulletin/850x646/san_francisco_airport_diagram_with_take_off_and_landing_dire ction_in_the_west_086e89f5fee8f2f655045082435540f1a05dda26.p ng

It's really a tribute to the industry and to the skills and professionalism of pilots and ATCOs that almost all of the millions of us who sometimes fly in and out of the Bay Area (and other busy destinations), in any seats, are alive and mostly undamaged.

[RetiredBA/BY]

That’s not debate. That’s just telling me to mind my own business.

Are you trying to tell me that at the end of long sectors pilots are always tired and, as a result, are not able to perform a safe approach?

If this is the case then why do the rules allow it?

I suppose the greatest irony here is that, in a Hawk, I completed an approach to 28R at SFO at the end of a 12 hour work day in 2012. So you see, it wasn’t in a wide body, but I have actually been there. I was tired too.

I realise you have a lot of experience in the commercial world (whereas all of my experience is fast jet) and I don’t but I do fly as a long haul passenger quite regularly. If the ability to fly a safe approach is truly so compromised by fatigue then surely something needs to be done about it.

BV


Most certainly not telling you to mind your own business, but I feel your undoubted military experience and expertise, without long haul airline experience might not allow you to have a balanced, informed debate !
Fatigue is a fact of life in such ops.and, I believe, always will be so long as humans are involved, particularly when airlines are trying to get as much productivity from the pilots as the law allows. Something the military does not do, so far as I know !

( I guess the short haul guys working to EASA limits are getting pretty tired too ! )

[JW411]

Retired BA/BY:

"Balanced/informed debate".

"You may have been fatigued BUT do you have any experience of long, often night, flights with big time changes and your circadian rythm shot to pieces."

"My worst, LHR to NBO, via CDG........."

Can you just remind us what the time shift is between London and Nairobi. My experience was that north/south was easy but east/west could just be a bit more difficult.

[RetiredBA/BY]

Only about two hours, and yes, east - West is a lot worse. That said, trying to sleep before a long night was not always easy, particularly in summer.
...but fatigue, not just tiredness, is not just about one trip, its the accumulation of repeated difficult flights.
Looking back at my career I remember that as the worst night’s work but perhaps time dims the memory !

I see fatigue is being discussed again in the BA pilots strike thread.

[cappt]

SFO is going down to one runway (landing) operation for the next twenty days due to the main runway 28L sinking in the middle at the intersection of runways one left/right. This is the runway they just did an overlay on two years ago.
Expect three hour delays and cancellations over the next month.
https://www.fly.faa.gov/flyfaa/flyfaaindex.jsp?ARPT=SFO&p=0 (https://www.fly.faa.gov/flyfaa/flyfaaindex.jsp?ARPT=SFO&p=0)
https://www.flysfo.com/media/press-releases/reconstruction-sfo-runway-28l-ready-begin-saturday-september-7th

[Sea Eggs]

The PNF was probably a training captain (Canadian eh?) paired with a low time Adelaide trained right seater trying to follow the magenta line.

[bud leon]

In my opinion the problem here largely lies with the controller not providing more information on where he was trying to position Cathay in the first place, and then continuing with the instruction when it was clear there was a risk of loss of separation.

That is, when he gave the first turn instruction he should have advised CPA 892 that the turn should be immediate to get in front of UAL 1515, and once that didn't happen, he should have lined CPA 892 behind UAL 1515. The direction to turn was quite late given the position of the aircraft in any case. I think it's very odd that he didn't change the sequence when it was clear the initial instructions were not followed and so could cause loss of separation. it seems reasonably clear that the controller was fixated on managing congestion, which he probably does his entire shift. I would say that is the root cause. Both pilots express a surrendering of control to ATC.

[ATC Watcher]

Both pilots express a surrendering of control to ATC.
And what else can you do when being vectored in Approach in a large busy airport ?

[787PIC]

Everybody is an "aviation expert" until the Lavatory backs up! :-)
I hear all you experts analyzing the RA maneuver, or the lack of by the United 'Guppy.'
At airports like SFO, with parallel visual approaches in progress, but still under positive radar control and both aircraft visually flying in formation, United pilots are allowed to place the transponder in TA. This is to prevent nuisance RA's in short final.
Unfortunately, the Cathay crew could not fly or follow the ATC instructions. In this case, I would have abandoned the approach the minute I see the other aircraft uncomfortably coming too close to mine. Way before the Tower finally waves the errant pilot off.
SFO/NorCal controllers are among the best in the business, given the incredible task in hand. They also assume a pilot of a state of the art A-350 flying into the San Francisco International Airport can follow basic ATC instructions and possess basic flying skills!

[misd-agin]

Retired BA/BY:

"Balanced/informed debate".

"You may have been fatigued BUT do you have any experience of long, often night, flights with big time changes and your circadian rythm shot to pieces."

"My worst, LHR to NBO, via CDG........."

Can you just remind us what the time shift is between London and Nairobi. My experience was that north/south was easy but east/west could just be a bit more difficult.

North/south isn't 'easy' because the time zone shift is minor. In the U.S. the departures are frequently at 1900-0000 and you fly all night arriving the next morning. The pickup the following evening is around 1900 and you land back in the U.S. around 0500-0600. I've got more pictures of relief pilots falling asleep on Deep South flights than I do from any other type of flying. Most of us doing the flying think the Deep South flying is harder than the 'tough' east/west flying you mentioned.

[gasbag1]

E They also assume a pilot of a state of the art A-350 flying into the San Francisco International Airport can follow basic ATC instructions and possess basic flying skills!

New type to the fleet, so how much time on type might be a consideration. Work overload, trying to make an A-350 slow to minimum speed in a few miles. Don't know the aircraft but I remember when London ATC tried to use what they considered normal separation behind a 747's when the DC-10 first started flying to LHR. Didn't work out so well for the DC-10 with a go arounds occurring frequently until ATC learned of the higher min. approach speed of the DC-10.

[OldLurker]

... I remember when London ATC tried to use what they considered normal separation behind a 747's when the DC-10 first started flying to LHR. Didn't work out so well for the DC-10 with a go arounds occurring frequently until ATC learned of the higher min. approach speed of the DC-10.That may be an ongoing phenomenon at LHR. Visiting a friend in London below the approach to 27L recently, I was interested to see three or four 777s dragging in with gear down, about 10nm from threshold. I wasn't there particularly to watch the aircraft but I didn't notice any other type with gear down so far out. Maybe a similar separation issue?

[Topic drift]

[IGNITIONS]

CX think they are the best pilots lol sorry mate.

[cactusbusdrvr]

On the 777 into LHR you are 180 kts until they slow you to 160 to 4 DME. Going downhill ( on the glide slope) around 500,000 pounds landing weight it won’t slow down to or hold 180 at flaps 15 gear up. Your choices are flaps 20 gear up or flaps 15 gear down. Most of us were trained to put the gear down before you go to flaps 20, mainly because if you go flaps 25 before the gear is locked down you will get the config warning.

[OldLurker]

cactusbusdrvr – Thanks, interesting. BTW, looking back I see that it was from a 787 that a stowaway’s body fell (https://www.aviation24.be/airlines/kenya-airways/stowaway-falls-boeing-787-8-landing-london-heathrow/) only a few blocks away from where I was. He’d apparently hidden in the wheel well, so I guess other types also need to have gear down as early as 10nm out.

[flydive1]

However, do we actually know wether they didn't follow the RA? It could have been a simple "monitor V/S" that didn't require any correction to their flight path, so they may not have done anything wrong after all.

Well, they said they got a RA to climb

[Meester proach]

cactusbusdrvr – Thanks, interesting. BTW, looking back I see that it was from a 787 that a stowaway’s body fell (https://www.aviation24.be/airlines/kenya-airways/stowaway-falls-boeing-787-8-landing-london-heathrow/) only a few blocks away from where I was. He’d apparently hidden in the wheel well, so I guess other types also need to have gear down as early as 10nm out.


things is, places like that want 180kts on base, then instantly 160 on the GS. Unless I’ve second guessed and slowed before GS, I’ll need to Chuck the wheels down and F15-20 to lose the 20 kts. And I can guarantee when I do that they’ll say “ maintain 180 for now “.

As per fatigue , EASA limits are seen as aspirational targets by some. I still can’t believe someone thought they were a good idea

i defy anyone to feel fresh at the end of ANY longhaul flight .

[Meester proach]

Let’s calm down a second. I for one basically agree with you.

Not always, but yes, myself and others are saying that fatigue sometimes plays a part. We were all unhappy when EASA rules came in, but sadly not enough of us nor BALPA nor our CAA were prepared to refuse the new FTL rules. So here we are. I put in fatigue reports as many others do, but companies are continually screwing more and more hours out of fewer and fewer pilots because the passenger will usually only pay for the cheapest fight. In my company, we fly three crew on flight times over 10 hours or so, so at least we each get a couple of hours’ rest outside the cockpit. However, for us that rest is in a seat in the passenger compartment, with a curtain around it. You can still hear - and are kept awake by - passengers around you talking.

We have video briefings about certain airports but sometimes it can be your first time there, and no amount of words on a page or videos can fully prepare you for an American airport..............

This Cathay crew screwed up. They missed a turn. I don’t know why, but very probably it was fatigue or inadequate preparation or briefing. The Cathay pilot on the radio sounds stressed to me. Maybe none of the flight crew had been to SFO before.

Some Long-haul pilots are short on manual handling skills. They shouldn’t be, but many are. If you only get two landings a month and are encouraged by your airline and FDM to fly with the automatics then this can be the result. I have suggested several times on other threads how this might be addressed.

Bear in mind that airliners are mostly flown by average pilots - (and I include myself in that analysis). Fast jet pilots from the forces are much better pilots than most of us because they were the best 0.0001% (or whatever) selected from all applicants.

UK ATC is the best in the world in my humble opinion and 18 years’ commercial flying. Many US airports would benefit by learning from UK trained and experienced ATCers such as those from Heathrow, Gatwick or Swanwick.

Fly safe :ok:






2 landings a month.....I should coco.

Most long haul pilots are knackered by the time we get anywhere, I’d simply reduce the margins if i decided it would be a good idea to start doing visual stuff. It would not be good airmanship at a lot the busy places we go .

Also your contention that forces pilots are better....well your opening a can of worms there aren’t you .

[Obama57]

I'm too tired to do a visual. I must have an ILS with the autopilot on. Oh, I forgot, it's the controllers fault, or it's United's fault. Whatever... IMHO, if you can't execute a visual approach at a busy airport, I'm sorry, you're not really an airline pilot.

[Bleve]

... if you can't execute a visual approach at a busy airport, I'm sorry, you're not really an airline pilot.

Just because you can execute a visual approach, doesn't mean you should. Risk management (aka airmanship) would suggest that if operating into an unfamiliar airport when fatigued after a longhaul flight, accepting and flying a visual approach might not be the wisest choice. To quote Frank Borman (Apollo 8 Commander):
A superior pilot uses his superior judgment to avoid situations which require the use of his superior skill.

[ScepticalOptomist]

I'm too tired to do a visual.............IMHO, if you can't execute a visual approach at a busy airport, I'm sorry, you're not really an airline pilot.

In a perfect world I’d agree. In the world of ultra long haul operations - it’s the reality. Not enough sectors - recency is an issue, as is genuine fatigue.

In my world the shortest sector we do is about a 15hr tour of duty back of the clock. I wish I was as good as when I was flying short domestic / international hops with multiple sectors a day. The reality is different.

Unless you have experienced it, I imagine it would be hard to empathise, but easy to criticise.

[hans brinker]

Just because you can execute a visual approach, doesn't mean you should. Risk management (aka airmanship) would suggest that if operating into an unfamiliar airport when fatigued after a longhaul flight, accepting and flying a visual approach might not be the wisest choice. To quote Frank Borman (Apollo 8 Commander):
Absolutely. Still, if cleared for the visual you are are either able or you tell the controller "unable" and get vectors for the ILS. Cathay accepted and screwed up, United got an RA (they might have not followed), the controller sat and watched it happen...

[Chris2303]

the controller sat and watched it happen...

I find it difficult to believe that the controller was sitting at his scope watching just those two airplanes (CX and UA) when he had others on his frequency.

[ATC Watcher]

Absolutely. Still, if cleared for the visual you are are either able or you tell the controller "unable" and get vectors for the ILS. Cathay accepted and screwed up,
Absolutely ! that is what we controllers do expect . Sadly that word seems to have vanished with the new magenta generation..
the controller sat and watched it happen...
You really think that once a ( any) clearance is issued and accepted we just sit there watching those 2 aircraft only to see if they comply ? Time to go and spend half an hour in a busy approach or center.
Which incidentally bring me back of one of the sad consequence of 9/11 : the suppression of Familiarization flights in cockpits by ATC.. You have today a full generation of controllers who have never visited an airliner cockpit or even talked to an airlines crew... Their only airborne experience might have been a single ride in a 172..
That could explain some might tell you to make a S turns to increase distance on a A350...

[Meester proach]

I'm too tired to do a visual. I must have an ILS with the autopilot on. Oh, I forgot, it's the controllers fault, or it's United's fault. Whatever... IMHO, if you can't execute a visual approach at a busy airport, I'm sorry, you're not really an airline pilot.


Ohhh , you are my hero.

Its not a case of can’t, it’s a case of “ is it a good idea ?” . That’s airmanship and risk management.

Yanks love their visuals , but it’s always the shorthaul ones doing it who’ve probably only come from Big falls, odawaydahoo.

[cappt]

Yanks love their visuals , but it’s always the shorthaul ones doing it who’ve probably only come from Big falls, odawaydahoo.

BigFalls for the fifth landing of the day in the same crappy weather system, what's your point? That's not fatigue?
Follow vectors to final and complete the visual is not asking a lot, heck you can even leave the autopilot on and let it do the approach, your call.

[FlightDetent]

BigFalls for the fifth landing of the day in the same crappy weather system, what's your point? That's not fatigue? ;) That's proficiency, not fatigue.

[gasbag1]

Any A-350 pilots out there that can tell us, how many NM does take to go from minimum clean speed to normal approach speed on the A-350?

[Veruka Salt]

Not sure if the incident aircraft was a 359 or 35K. 359 slows down like a 744, 35K is more slippery, like an A330. Commencement of speed reduction from 250 kts to be CONF 2/180/3000 at 10 nm starts around 20nm for a - 900, 23nm for a - 1000. Loose, ballpark figures.

[DaveReidUK]

Not sure if the incident aircraft was a 359 or 35K.

Not helped by the fact that after nearly 100 posts, we still don't seem to know when the incident happened.

[Veruka Salt]

None of us in the airline know anything about it either. Unusual, as they are generally very good about promptly passing info. regarding operational events.

[billyt]

I'm too tired to do a visual. I must have an ILS with the autopilot on. Oh, I forgot, it's the controllers fault, or it's United's fault. Whatever... IMHO, if you can't execute a visual approach at a busy airport, I'm sorry, you're not really an airline pilot.

How much long haul flying do you do?

[glofish]

Here: 40+ years of airline flying, thereof 24 on LH and ULH.
byllit, your answer is bs.
I almost fully agree with Obama57, although i would go further and say that if you can't execute a visual at a busy airport with marginal weather: Stay away from being an airline pilot! That's exactly one of the reasons you are still in the cockpit for!
You will not do that on every second flight, but you need to be able to do it and therefore you need to practice it every now and then.
I see too many young colleagues turn pale when i even only raise the idea of a visual. Brave new world!

[deja vu]

Is a 14 hour sector with 6-8 hours in the bunk still considered long haul?

[ScepticalOptomist]

Is a 14 hour sector with 6-8 hours in the bunk still considered long haul?

Departing at 2030 local for a 14hr flight, I’d say is long haul,

Even with about 6.hrs “off duty” in the bunk that means that two pilots are trying to have a snooze between 2130-0030 and then again about 0330-0630 body time.
The other two pilots are taking their rest at 0030-0330 and then 0630-0930 body time.

Assuming they can go straight to sleep and don’t get disturbed by noise / turbulence / their bladders etc they would feel about average when they all turned up prior to descent.

I average about 2hrs of broken sleep at each break and feel ok by ToD. Not at 100% but good enough. Let’s not pretend that we’re as rested as if we’d been tucked up in our beds / hotel for 6hrs straight.

Having also spent half of my career doing multi sector domestic ops I can categorically say that long haul ops hurt way more - for me. That is offset by more days at home between trips etc but the overall effect on my body and flying skills is that I find long haul tougher overall than 2 man domestic ops day in day out.

Everyone is different.

[PukinDog]

Ohhh , you are my hero.

Its not a case of can’t, it’s a case of “ is it a good idea ?” . That’s airmanship and risk management.

Yanks love their visuals , but it’s always the shorthaul ones doing it who’ve probably only come from Big falls, odawaydahoo.







Predictable anti-Yank snark doesn't change the fact that this Cathay crew had a loss of SA, was well behind their own aircraft, and didn't comply with clearances/instructions they acknowledged. The failure of airmanship and risk management occurred long before the offer of a visual approach to try and salvage the mess was made or the RAs occurred.

Cathay was issued a visual approach for 28L (4:00 minutes into the recording) after they had already stuffed-up the clearance to intercept and capture the electronic guidance/localizer for 28L, a stuff-up that resulted in a loss of separation with UAL on approach to the parallel 28R, and only after Cathay was asked if they wanted to proceed using one. The reason for controller's visual approach offer was, since Cathay didn't follow their original clearance to join the 28L localizer, because it the only option remaining that would allow Cathay to maneuver back to the left for 28L. The offer and question...not the clearance...for a 28L visual approach came at 3:45, after which Cathay responded they could indeed continue using one. Only then (at 4:00), was the visual approach clearance issued to Cathay accompanied by the controller's re-iteration that they were to proceed to 28L and UAL was on 28R.

To review what led up to it.

Cathay was being vectored to 28L and given a heading of 010 which they acknowledged but didn't execute. When finally prompted by the controller to verify they were turning, Cathay asked for the controller to repeat the heading. The controller then asked them for a "sharp turn" (due to not following the previous vector) to the new heading of 330. At 1:32 Cathay was then given a heading of 310 and clearance to intercept the localizer for 28L, which Cathay acknowledged and read back correctly at 1:35. However, instead of following what they were cleared-for and read back, they proceeded to blow through the 28L localizer and encroached on the approach path for 28R.

Prior to that and before where the localizer intercept was to occur, the UAL traffic was pointed out to Cathay by the controller which Cathay acknowledged they had visually. Cathay was told that that UAL traffic was for 28R and instructed to maintain visual separation from them, which is a standard ATC instruction in VMC conditions when approach paths are in close proximity. At 1:58 the controller first pointed out the UAL traffic on approach stating it was on the approach for 28R. Cathay acknowledged they had them visually at 2:05 and, after a repeat of the subsequent instruction, at 2:23 read back/acknowledged the instruction to maintain visual separation. However, they obviously failed to comply.

At 1:35 Cathay was issued and acknowledged the final vector and clearance to intercept 28L localizer. At 2:25 while on that heading, they acknowledged visual contract with traffic for the parallel runway and the instruction to maintain visual separation. The failure to intercept the localizer/final approach course for 28L and encroachment into 28R's occurred 40 seconds later at about 3:05.

The real question(s) is why didn't the Cathay crew comply with simple ATC vectors to intercept the localizer for 28L and, after acknowledging they had visual contact with traffic on a parallel runway (28R) and would maintain visual separation, blew through their final approach course to wind up underneath that other traffic lined-up for 28R. It's as if they thought they were landing on 28R, tuned the wrong freq, and never looked out the window again.

All the babble of visual approaches (initially when offered in this case a simple side-step while still above 2,000') being undoable, the supposed superiority of UK controllers over US controllers, the "dangerousness" of SFO, what UAL should have done, ULH willy-waving, etc etc completely misses the point. One would think on a professional pilot forum that pilots at least would be able to recognize where the snowball began to roll downhill and address/discuss the possible how and why it did. We're left with only possibilities since with no CVR leading up to the breakdown we can only suppose what was briefed and expected and to what degree by the crew long before this recording begins, but obviously it went to hell.

[tomuchwork]

So many parallel approaches in that big world we are all living in. Many of them in very busy environment. Parallel approaches require very good crew coodination AND skills and it seems that the Cathay messed up big time here. Replying to ATC is one thing, DOING it a totally different thing apparently.

Blaming the United is just cheap. Nothing more to say. Do not know their procedures, in my outfit you get most likely fired for not following an TCAS RA. But the one who clearly messed up is the Cathay. Congrats. The world is getting safer and safer(and the pilots worse and worse) ^^.

P.S.: I LIKE my visuals as well and I fly for a bad loco over here in Europe that does not support visuals(because they are very very dangerous apparently).

PPS: Some blame the long haul for this errors. Well. I have done my very good share of long haul command in 2 pilots ONLY ops for one of the big 3 in the ME some years ago. Some try to excuse the crew by saying they are tired but then, on the other hand, are they not getting paid(and they should be qualified) to handle the threats? You know you are going to fly into a very busy environment, you know you will be tired, you know(if you are somewhat experienced) what to expect - THAT's WHAT THEY PAY YOU FOR(and even nowadays this is not really peanuts what they pay you). Pushing the "engage AP" button is not what we are all paid for, I assume that should be a common understanding. It is handling difficult situations and managing to end up without a report required to be filled out.

[PukinDog]

In my opinion the problem here largely lies with the controller not providing more information on where he was trying to position Cathay in the first place, and then continuing with the instruction when it was clear there was a risk of loss of separation.

That is, when he gave the first turn instruction he should have advised CPA 892 that the turn should be immediate to get in front of UAL 1515, and once that didn't happen, he should have lined CPA 892 behind UAL 1515. The direction to turn was quite late given the position of the aircraft in any case. I think it's very odd that he didn't change the sequence when it was clear the initial instructions were not followed and so could cause loss of separation. it seems reasonably clear that the controller was fixated on managing congestion, which he probably does his entire shift. I would say that is the root cause. Both pilots express a surrendering of control to ATC.

The controller was perfectly clear and they weren't being sequenced for the same runway. Cathay was being vectored and was cleared to intercept the localizer for 28L. On the other hand, UAL 1515 (the traffic pointed out by the controller which Cathay acknowledged they had visually and instructed to maintain visual separation from, also acknowledged) was on approach for 28R. In the recording the Left and Right runway distinctions are made and Cathay was advised after visual contact that UAL 1515 was going to the Right; they didn't have to infer, deduce, or suppose it. If you listened to the recording I'm surprised you missed it.

Cathay would have known before this recording began that they were being vectored/positioned for 28L, something that's clear in the instructions from ATC that we do hear in the video and is always stated by the Approach controller to the crew when initiating vectors off the published arrival, something that occurred before the recording began. The Cathay crew would also know there were parallel runway ops to 28R in progress. ATIS states it and 2 the daisy-chains of aircraft on TCAS makes it obvious.

28L and 28R are only 750' apart at SFO, but with a final vector of 310 to intercept the 28L localizer inbound course of 284, that's only a 26 degree intercept angle. Hardly something difficult for automatics or by hand, it's downright mild, and raises the question of correct freq tuned or automation that went unarmed.

The problem wasn't an ATC controller fixating on his original sequencing plan, it was Cathay crew fixating on something else besides intercepting the correct localizer as instructed/cleared/acknowledged and maintaining the lateral visual separation of UAL 1515 on approach for the parallel runway they said they would maintain.

Contrary to the cause being both pilots surrendering control to ATC, how about a crew not following clear ATC instructions/clearances they've acknowledged, read back, and accepted?

[Uplinker]

................The real question(s) is why didn't the Cathay crew comply with simple ATC vectors to intercept the localizer for 28L and, after acknowledging they had visual contact with traffic on a parallel runway (28R) and would maintain visual separation, blew through their final approach course to wind up underneath that other traffic lined-up for 28R. It's as if they thought they were landing on 28R, tuned the wrong freq, and never looked out the window again...........

(my bold)

Great post, PukinDog: This seems to be the most likely reason behind the confusion on the Cathay flight deck. Agree with the rest of your (unquoted) post #102 too :ok:

[tiddles52]

is this near miss being investigated at all by the authorities,NOT so anybody can get into trouble, but so lessons can be learnt or at the very last a cold cool professional analysis done of all the causal events, so people can't say years down the line they never knew.

This is the backbone of aviation safety, using every little thing as a no-fault learning exercise and slowly closing the holes in the Swiss Cheese.