yyztpa wrote:

http://avherald.com/h?article=4ab79f58
"It is estimated that AC-759 overflew the first two aircraft by 100 feet, the third by about 200 feet and the last by 300 feet. The closest lateral proximity between AC-759 and one of the aircraft on taxiway C was 29 feet."

If this was frightening before, it just got a lot worse. I had assumed that the pilots would figure it out, realize that they were looking at aircraft in their path, and perform a go around later... and that they were only "beaten to the punch" by ATC (others posting had similar comments).

But according to the article they were already over the taxiway when they initiated a go around..., and they came very (VERY) close. It suggests to me that they would not have caught it in time, but for the ATC wave off.

I posted this question earlier, but received no response, so I'll try again (I'm hoping that someone with actual flight experience can help): At what point is it necessary to perform a missed approach? When the crew of AC759 saw lights on the runway, less than a mile out, wouldn't that be the time to go around and figure it out from a safe height? OR, when the tower assures that there are no aircraft on the runway, but their eyes tell them otherwise?

Thanks in advance!

PS - I'm flying AC YYZ-SFO in a few months, so this is way too close for comfort.

from the avherald link:
The vertical miss was 100'. That's 7 seconds to impact if they hadn't been told to go around (800 fpm descent on 3 degree glidepath). You are very generous to allow that the pilots would have detected their error in time.

We'll never know if they would have reacted in those 7 seconds I guess.

DaufuskieGuy wrote:

You are very generous to allow that the pilots would have detected their error in time.

Well considering they already thought something was amiss, I'm not sure it is generous at all to think they would not have gone around on their own.

One of the contributing factors making SFO so much in need of a foolproof landing system is that so many aircraft are on approach with dead-tired flight crews after very long flights. Lindbergh Field in San Diego used to be a poster child, but really SFO has so many intersecting factors - fog, approach over water, tired crews, inconsistent runway lighting and use, language barriers due to a lot of international flights ... This had better be a test case that leads to proactive improvement. Why wait until the accident - the writing is on the wall for all to see...

DaufuskieGuy wrote:

from the avherald link:
The vertical miss was 100'. That's 7 seconds to impact if they hadn't been told to go around (800 fpm descent on 3 degree glidepath). You are very generous to allow that the pilots would have detected their error in time.

We'll never know if they would have reacted in those 7 seconds I guess.

I think they would have gone around. A lot comes into view in the last 7 seconds at 140kts. It would have had to have been a no-delay-on-the-pitch-up type of go-around but still I think they'd notice the beacons and once their own landing lights illuminate the first jet, they'd be on the go.

CO953 wrote:

The airport runway/taxiway lighting schema should appear EXACTLY THE SAME to each approaching pilot, EVERY SINGLE TIME, identifying each and every runway and taxiway, until it is so thoroughly drummed into the pilot's head by repetition that he/she knows exactly the number and layout of runways and taxiways, even with no lights at all. This business of having two runways, 28R and 28L, plus a parallel taxiway, then shutting off the lights on one of them and shutting off the X at night, forcing pilots to decide whether he's looking at 28L and 28R, or 28R and a taxiway, seems an insane and unnecessary shell game that repeatedly endangers lives.

It would appear some UI standardization would be helpful.

goboeing wrote:

DaufuskieGuy wrote:

from the avherald link:
The vertical miss was 100'. That's 7 seconds to impact if they hadn't been told to go around (800 fpm descent on 3 degree glidepath). You are very generous to allow that the pilots would have detected their error in time.

We'll never know if they would have reacted in those 7 seconds I guess.

I think they would have gone around. A lot comes into view in the last 7 seconds at 140kts. It would have had to have been a no-delay-on-the-pitch-up type of go-around but still I think they'd notice the beacons and once their own landing lights illuminate the first jet, they'd be on the go.

How long do the engines take to spool up and start providing meaningful thrust?

Whiteguy wrote:

hayzel777 wrote:

I thought foreign airlines were banned from visual approaches after the Asiana crash and EVA Air near miss? Was that ban removed?

Visual approaches were never banned.....

What Eva Air near miss?

My bad, they were banned while the ILS was under construction at SFO but the ban was lifted after it was completed.

https://www.flightglobal.com/news/artic ... rt-388850/

The same month Asiana crashed, EVA 28 was on visual when the tower noticed it was way too low and initiated a go around. Had the tower not noticed, we would've potentially had a second crash.
http://avherald.com/h?article=465e38db

My understanding is that separation is based on transponder reported altitudes, so the actual separation of, let's say 1000 feet is not a clear 1000 feet between all parts of an aircraft, but closer to 1000 ft from their centres (therefore 1000 ft separation between two A380s would be less separation that 1000 ft between two A320s).

AVHerald is reporting the estimated vertical separation as 100ft, which is not a lot.. I wonder what the true "gap" between the AC and the first aircraft of taxiway C was.

I have been following this thread with great interest and what is abundantly clear is that a MAJOR accident could have happened, but didn't. This is absolutely the time to evaluate the situation and contributing factors just as if it had happened. What I can say is that it seems the emphasis on situational awareness is paying dividend. Regardless of who should get credit for averting the accident, there were certainly many people who were aware of their situation and were willing to speak out when something didn't seem right. Kudos to all!

CO953: "As a non-pilot, I am still dumbfounded at how something so simple can be treated as so complicated by the industry."

Frustrating, isn't it? Regarding your (relevant) rant about the thrust levers, it's quite obvious that the brain is very good at recognizing patterns and if you mess the lighting pattern of runways up from time to time, the possibility of a mistake increases, that is an obvious fact, nothing else. How much the risks increase is another matter. I think all the people screaming about how bad the Airbus crew acted needs to reconsider and put themselves in their seats. A lot of things influences the approach, and if I'm not mistaken, the "Quiet Bridge Visual" approach may have been used here, meaning a last-minute turn after San Mateo bridge in order to line up to the runway. Your red X'es at the beginning of every taxiway should be a no-brainer.

Regarding spool-up times, the normal procedure on landings is to remain on a higher thrust than you'd think was needed (and hence, high RPM) to minimize just that. You can often hear the engines spinning at rather high RPMs during landings. To help reduce speed and altitude, spoilers and flaps are used. As a result, you can fairly instantly get more thrust and in seconds, the flaps can be moved from full deployment (step 3 in Airbus?), which causes a lot of drag, to a lower setting, almost maintaining the same lift but reducing drag significantly. Then again, you won't go sky-rocketing in under a second, but perhaps 2-3 seconds would give you plenty of thrust. I'd guess you could always level off with the thrust and drag you have and would be able to sustain level flight until thrust is increased and flaps are retracted one notch (basically to a normal take-off setting since you'd still need them to produce the extra lift in this low-speed scenario). I'm sure there are pilots here that can explain this a lot better. I wonder how the thrust from the A320 felt in those four airliners lined up at Charlie...

Regarding the pilots reacting on their own, I think it's quite possible that the AC crew realized their mistake during the conversation with ATC and were in the process of initiating a go-around as they heard ATC order it, but we won't be able to determine that before we see the flight recorder data or other info.

/Fredrik

YYZYYT wrote:

I posted this question earlier, but received no response, so I'll try again (I'm hoping that someone with actual flight experience can help): At what point is it necessary to perform a missed approach? When the crew of AC759 saw lights on the runway, less than a mile out,

There is no rule. As long as you think the runway will be clear by the time you land. At some busy airports it is not uncommon to be coming over the threshold as the previous departure rotates or the tail of the previous arrival clears.

https://goo.gl/images/YsrxwN

All of the taxing airplanes cued up to take off would only have had their beacons and nav lights on. Those lights all easily blend with all the other lights from the runways, taxiways, nearby roadways, service roads, etc.

There may be issues of a weakness in CRM, perhaps some additional training with refreshers for all pilots (and not just for AC) as to approaches to SFO should be put in. I agree that ILS should be used as an aid, so long is properly calculated and used.

Prior to TOD there should have been a discussion of the NOTAM indicating KSFO Rwy28L out of service and the associated approach situational awareness. One would therefore have expected takeoff traffic (based on the METAR winds and ATIS) to be using 28R for takeoffs as well as landings. As such it would not have been surprising for a few aircraft to be backed up on taxiway C for 28R takeoff.

The 28R FMS bridge visual would have been selected and set in the FMC. This custom/company generated and approved FMS based approach has all the waypoints and altitude crossing restrictions included. They entered the 28R FMS bridge visual at TRDOW WP (with ATC approval) and should have been in PFD modes LNAV/VNAV and A/T and A/P engaged.

Upon entry (and with necessary MCP altitude change) the aircraft automatically follows the FMS approach waypoints and altitude restrictions just like an RNAV approach. Upon FMS visual exit WP (around 4 NM from threshold) and prior to minimums they would initiate a manual landing (A/P and A/T disengaged). Knowing that 28L is closed and with a clear night (per the METAR) the approach runway lighting of 28R would have been glaring including the centerline lights, edge lights, runway end lights, touchdown zone lights, lead in lights, PAPI, etc. And the path to 28R threshold would have been clear on the ND and F/D guidance.

But if one is looking at green taxiway guidance lighting and an absence of the runway approach lighting, etc. one should know they’re about to land on a taxiway. There was obviously crew confusion, as they asked ATC about seeing lights (aircraft) on the runway (which were actually on taxiway C)…at that point they should have initiated a (early) go around themselves, and not ended up overflying taxiway C by 0.25 NM and still descending, before being instructed (thank God) by ATC to go around.

This was a very serious incident and potentially catastrophic due to the reported actual conflict distances and which could have involved several large aircraft on taxiway C. This incident should not be taken lightly. All just my opinion.

A different thread on this forum says:

"From Air Disaster Volume 3:

CFM56-5:
29% N1 to 67%: 4 seconds
67% N1 to 83%: 1 second"

And also that there is an FAR that specifies a maximum time:

"There is an industry standard. They must accel from flight idle to 95% rated thrust in 5 seconds or less. It is found in FAR Section 33.73(d)."

treetreeseven wrote:

How long do the engines take to spool up and start providing meaningful thrust?

Post-overhaul engine testing requirements usually have a maximum spool-up time of 6-7 seconds (that is a 'snap accel' from approach idle to ~90% of takeoff power). This varies slightly between GE/PW/RR, but in general, it's in this ballpark..

Longhornmaniac wrote:

As a pilot, I would absolutely hate a "big freaking brilliant eye-piercing king-kong sized red "X," since I want to actually be able to see.

Make it light up only when it detects an approaching airplane at a distance far enough for a go-around. You know, kind of like when you walk to someone's front door and their motion sensor light comes on. It can't be that difficult, or expensive.

treetreeseven wrote:

How long do the engines take to spool up and start providing meaningful thrust?

Good question, but I don't think it would have been a factor at all.

From ground idle to 50% is a little sluggish...the take a few seconds to [unevenly] spool up to hit that mark. But from 50% to higher thrust is very quick, almost uncomfortable quick in the air when you hear the engines surge to meet the demands of a relatively small speed adjustment request by the pilot.

Reiterating what I had posted above, I think there is no way they ever would have hit the airplanes on Charlie.

Once their own landing lights were showing a jet right in front of them they'd be hauling back on the sidestick and climbing out.

We practice rejected landings from less than 50' AGL in the sim, it's quite responsive. Not saying 50' wouldn't put them through the top of a vertical stabilizer but I think they'd be climbing out at that point, above them.

FlyBigDeltaJets wrote:

treetreeseven wrote:

How long do the engines take to spool up and start providing meaningful thrust?

Post-overhaul engine testing requirements usually have a maximum spool-up time of 6-7 seconds (that is a 'snap accel' from approach idle to ~90% of takeoff power). This varies slightly between GE/PW/RR, but in general, it's in this ballpark..

But no way were they at idle that late in the approach.

You can't maintain a 3 degree glideslope fully configured at anything near Vref with idle power (Asiana anyone?).

The engines were already spooled and GA power is nearly instant.

barney captain wrote:

FlyBigDeltaJets wrote:

treetreeseven wrote:

How long do the engines take to spool up and start providing meaningful thrust?

Post-overhaul engine testing requirements usually have a maximum spool-up time of 6-7 seconds (that is a 'snap accel' from approach idle to ~90% of takeoff power). This varies slightly between GE/PW/RR, but in general, it's in this ballpark..

But no way were they at idle that late in the approach.

You can't maintain a 3 degree glideslope fully configured at anything near Vref with idle power (Asiana anyone?).

The engines were already spooled and GA power is nearly instant.

In an Airbus Idle is considered spooled due to approach idle EEC commands

mcdu wrote:

barney captain wrote:

FlyBigDeltaJets wrote:

Post-overhaul engine testing requirements usually have a maximum spool-up time of 6-7 seconds (that is a 'snap accel' from approach idle to ~90% of takeoff power). This varies slightly between GE/PW/RR, but in general, it's in this ballpark..

But no way were they at idle that late in the approach.

You can't maintain a 3 degree glideslope fully configured at anything near Vref with idle power (Asiana anyone?).

The engines were already spooled and GA power is nearly instant.

In an Airbus Idle is considered spooled due to approach idle EEC commands

Thanks for that - but again, it seems very unlikely these engines were at idle.

hayzel777 wrote:

Whiteguy wrote:

hayzel777 wrote:

I thought foreign airlines were banned from visual approaches after the Asiana crash and EVA Air near miss? Was that ban removed?

Visual approaches were never banned.....

What Eva Air near miss?

My bad, they were banned while the ILS was under construction at SFO but the ban was lifted after it was completed.

https://www.flightglobal.com/news/artic ... rt-388850/

The same month Asiana crashed, EVA 28 was on visual when the tower noticed it was way too low and initiated a go around. Had the tower not noticed, we would've potentially had a second crash.
http://avherald.com/h?article=465e38db

Actually after the Eva near miss after Asian 214, all foreign flag carriers at SFO were banned from doing visual approaches and parallel landings for about 2 months. British Airways and Lufthansa complained heavily and the decision was eventually reversed. The single landing requirement for foreign flag carriers also started causing delay problems as this would reduce the flow during the peak operating hours.

cat3appr50 wrote:

Prior to TOD there should have been a discussion of the NOTAM indicating KSFO Rwy28L out of service and the associated approach situational awareness. One would therefore have expected takeoff traffic (based on the METAR winds and ATIS) to be using 28R for takeoffs as well as landings. As such it would not have been surprising for a few aircraft to be backed up on taxiway C for 28R takeoff.

The 28R FMS bridge visual would have been selected and set in the FMC. This custom/company generated and approved FMS based approach has all the waypoints and altitude crossing restrictions included. They entered the 28R FMS bridge visual at TRDOW WP (with ATC approval) and should have been in PFD modes LNAV/VNAV and A/T and A/P engaged.

Upon entry (and with necessary MCP altitude change) the aircraft automatically follows the FMS approach waypoints and altitude restrictions just like an RNAV approach. Upon FMS visual exit WP (around 4 NM from threshold) and prior to minimums they would initiate a manual landing (A/P and A/T disengaged). Knowing that 28L is closed and with a clear night (per the METAR) the approach runway lighting of 28R would have been glaring including the centerline lights, edge lights, runway end lights, touchdown zone lights, lead in lights, PAPI, etc. And the path to 28R threshold would have been clear on the ND and F/D guidance.

But if one is looking at green taxiway guidance lighting and an absence of the runway approach lighting, etc. one should know they’re about to land on a taxiway. There was obviously crew confusion, as they asked ATC about seeing lights (aircraft) on the runway (which were actually on taxiway C)…at that point they should have initiated a (early) go around themselves, and not ended up overflying taxiway C by 0.25 NM and still descending, before being instructed (thank God) by ATC to go around.

This was a very serious incident and potentially catastrophic due to the reported actual conflict distances and which could have involved several large aircraft on taxiway C. This incident should not be taken lightly. All just my opinion.

So how "visual" was this visual approach? At what point were there eyes permanently outside looking for the runway?

good discussion...
.
"We learn from our mistakes..."

...and we learn from our near-misses as well...
.

barney captain wrote:

FlyBigDeltaJets wrote:

treetreeseven wrote:

How long do the engines take to spool up and start providing meaningful thrust?

Post-overhaul engine testing requirements usually have a maximum spool-up time of 6-7 seconds (that is a 'snap accel' from approach idle to ~90% of takeoff power). This varies slightly between GE/PW/RR, but in general, it's in this ballpark..

But no way were they at idle that late in the approach.

You can't maintain a 3 degree glideslope fully configured at anything near Vref with idle power (Asiana anyone?).

The engines were already spooled and GA power is nearly instant.

I don't doubt any of the points you made, but keep in mind approach/flight idle is different (higher) than ground idle. I was just laying out a "worst case" answer of spool-up time in response to the previous posters question. The engine testing requirements for snap accel would assume a worse-case scenario where GA is required after engines have been retarded to idle and the aircraft has touched down.

Spooling up the engines and getting full power is one thing. Accelerating the aircraft is another thing.

The engines would have been on spool, it would have been about 2 seconds to get to TOGA thrust. This is a non issue in the case.

Thinking outside the box here....

Is it so hard to have a GPS-driven heads-up display on the windshield that projects an overlay of the lights/landmarks that the pilots should be seeing?

CO953 wrote:

Thinking outside the box here....

Is it so hard to have a GPS-driven heads-up display on the windshield that projects an overlay of the lights/landmarks that the pilots should be seeing?

Some aircraft have HUDs.....it depends on the airframe and the airline. I'm not sure if AC have them or not

Well, this landing thing has to be sorted out, industry-wide. Aviation has become too precise and safe to have failing to identify the runway be the last primitive failure mode that kills pilots and passengers..... Think back how many commercial and civilian carriers have had short or misplaced "landings" in the last 10-15 years, and then compare that to the number of other-type accidents.

We're on the cusp of self-driving cars. Surely the aviation industry can tell a pilot if he is looking at a runway or a taxiway, and have some big automated STOP sign in his face as he's about to land on 1,000 people.

(This is a bit personal for me, as I lost a young friend in the Aspen, Colorado Gulfstream crash in 2001 due to failure to identify and acquire the runway.)

Once an aircraft is near the runway, modern technology should make it easy-peasy to identify it.

CO953 wrote:

Well, this landing thing has to be sorted out, industry-wide. Aviation has become too precise and safe to have failing to identify the runway be the last primitive failure mode that kills pilots and passengers..... Think back how many commercial and civilian carriers have had short "landings" in the last 10 years, and then compare that to the number of other-type accidents.

We're on the cusp of self-driving cars. Surely the aviation industry can tell a pilot if he is looking at a runway or a taxiway, and have some big automated STOP sign in his face as he's about to land on 1,000 people.

(This is a bit personal for me, as I lost a young friend in the Aspen, Colorado Gulfstream crash in 2001 due to failure to identify and acquire the runway.)

Once an aircraft is near the runway, modern technology should make it easy-peasy to identify it.

You have to look at the bigger picture. I understand your frustration but simply look at the amount of flights there are on a daily basis. Then look at how many land on taxiways. The number is 99.999999% safe. Not everything can be regulated. Humans make mistakes. If it wasn't for human error the aviation industry would have ZERO fatalities.

It is very easy for me to identify a runway.......the centerline lights aren't green and the edge lights are not blue......it doesn't get too much simpler than that. But if you are fatigued and tired who knows......

CO953 wrote:

(This is a bit personal for me, as I lost a young friend in the Aspen, Colorado Gulfstream crash in 2001 due to failure to identify and acquire the runway.)

That's not why that crash happened. It happened because of the crew descending below minimums without adequate visual reference, though the overarching cause was pressure from the passengers to land at Aspen instead of diverting.

Exactly. We are saying the same thing in different languages. At Aspen they never confirmed acquisition of the runway. Read the transcript.... there was never an agreed affirmative. Yes, the charter customer in the cockpit made it worse. The controller saw the jet come out of the twilight haze pointed directly at the cliff and hit the crash button. The pilots committed to the landing and never between them agreed that they saw the runway.\

Thank God that the Air Canada crew had some visibility. Otherwise at SFO we have the worst crash in world history, potentially killing 1K people

Try this again at SFO with dense fog and you have five piles of ash on the taxiway.

CO953 wrote:

Exactly. We are saying the same thing in different languages. At Aspen they never confirmed acquisition of the runway. Read the transcript.... there was never an agreed affirmative. Yes, the charter customer in the cockpit made it worse. The controller saw the jet come out of the twilight haze pointed directly at the cliff and hit the crash button. The pilots committed to the landing and never between them agreed that they saw the runway.\

Thank God that the Air Canada crew had some visibility. Otherwise at SFO we have the worst crash in world history, potentially killing 1K people

Try this again at SFO with dense fog and you have five piles of ash on the taxiway.

We are not saying the same thing. A decision to descend below minimums on an instrument approach is not the same as an inability to distinguish the runway under visual conditions. They are two very different forms of error.

And as has been mentioned multiple times in this thread, if there had been dense fog in SFO this would not have happened.

Mir wrote:

CO953 wrote:

Exactly. We are saying the same thing in different languages. At Aspen they never confirmed acquisition of the runway. Read the transcript.... there was never an agreed affirmative. Yes, the charter customer in the cockpit made it worse. The controller saw the jet come out of the twilight haze pointed directly at the cliff and hit the crash button. The pilots committed to the landing and never between them agreed that they saw the runway.\

Thank God that the Air Canada crew had some visibility. Otherwise at SFO we have the worst crash in world history, potentially killing 1K people

Try this again at SFO with dense fog and you have five piles of ash on the taxiway.

We are not saying the same thing. A decision to descend below minimums on an instrument approach is not the same as an inability to distinguish the runway under visual conditions. They are two very different forms of error.

And as has been mentioned multiple times in this thread, if there had been dense fog in SFO this would not have happened.

Thanks for the response, and I defer to your technically correct description of the Aspen error. As a non-pilot, I can see what you are saying technically. But I have read that transcript many times over the years and just now I read it again and I still cannot see any consensus between the two pilots that the runway was in sight.

Maybe this is what I'm trying to say. I am a non-pilot but a dedicated student of aviation who has spent 40 years reading accident transcripts.

If we back away from the technical accuracy well trained into industry personnel, and hew closer to a non-jargon explanation, we still see that the Aspen jet and the Air Canada jet were committed to landing but did not truly know the touchdown point. Spoken from a layman's view, the Avjet pilots at Aspen in 2001 firewalled the throttles at the last moment once they saw the runway and tried but failed. The Air Canada pilots at SFO did the same thing and survived.

There but the grace of God.

In my opinion, as the layman trying to provide some outside opinion to aviation professionals who have the training I don't, but maybe sometimes get too close to the subject to call a spade a spade:

The error is getting that close to the runway and not knowing where the hell you're going.

Unacceptable. Totally. And I hugely blame the airport with their retarded lights policy.

The deaths at Aspen in 2001 led to further refinement of CRM procedures with regard to charter customers. I'm hoping that the near-tragedy at SFO can lead to drastic changes in runway lighting management, BEFORE the sacrificial deaths that are usually required before the industry will stir itself to fix a glaring weakness.

CO953 wrote:

Thinking outside the box here....

Is it so hard to have a GPS-driven heads-up display on the windshield that projects an overlay of the lights/landmarks that the pilots should be seeing?

A few years back I got to fly a sim with the Rockwell Collins Synthetic Vision system. It is unbelievably real looking, just imagine everything bathed in highlighter green at night/in weather. The Rockwell folks put everything you can imagine to reduce visibility and I had no trouble finding the runway and staying away from mountains/obstacles. You literally are looking at things outside as if it's a perfectly clear day.

No system is foolproof, and no pilot is perfect, but these things can only help increase situational awareness IMO.

Glad no metal got bent. A situation like this is pretty alarming. If I was the pilot on the first UA plane, I'd need new underwear!

Thanks to the experts here for enlightening me on the procedures SFO typically employs regarding shutting off runway lights and disabling the approach strobes. I'm a aviation enthusiast, but not a pilot, so it is interesting to hear how pilots with experience shooting approaches into SFO view this incident over the sometimes inaccurate media reports. Certainly makes it easier to understand how the AC759 crew could have made this mistake as I was having a hard time understanding how the crew could mistake the taxiway for 28R given the weather conditions and my mistaken assumption that 28L was illuminated.

I can't help think of this incident being another example of why SFO needs to fill in a part of the Bay to build another parallel runway off the north end of 19L/R. It would definitely help during the winter months when the airport sees IFR conditions more often. Unfortunately local politics makes that a non-starter. Sadly, I dont see that changing even if a similar incident were to occur with a less ideal outcome.

I hate doing this on airliners because no one likes their exuberance dismissed, but lets look at what we have:

A doctored recording of the ATC conversation as the only evidence. A retired "aviation expert" saying this could have been worse than the Canary Islands. So this news report that is the primary source of information is clearly showing a sensationalist bias. Aside from that report there is no evidence indicating that anyone was in any danger at any time. In fact there is no evidence to indicate that PIC made an error of judgment worth including in his record. So this whole event is as yawn worthy as it is disaster averted. At least wait for there to be some blood in the water before people are condemned.

There must be video. Anybody seen a link?

CO953 wrote:

Thank God that the Air Canada crew had some visibility. Otherwise at SFO we have the worst crash in world history, potentially killing 1K people

Try this again at SFO with dense fog and you have five piles of ash on the taxiway.

If SFO had dense fog Air Canada would have followed the ILS all the way down most likely on Auto Pilot and would have broke out directly in line with the runway.

Replying to hivue’s post #122 above.
As I noted the 28R FMS bridge visual is essentially a VNAV/LNAV approach and very similar to an RNAV approach requiring the proper scan of the ND, PFD, etc. and outside. I would have flown this FMS approach via LNAV/VNAV, A/P, and A/T. Once turning to the 28R FMS bridge visual entry at TRDOW I would be scanning the PFD, ND, etc. as well as out the windshield. ATC is eventually going to ask you if you have KSFO or the bridge in sight before TRDOW. Upon reaching the FMS bridge visual exit at F101D (just after the SM bridge) it is obviously essential to be primarily looking out the windshield, as one would be transitioning to a manual landing on 28R (with F/D guidance available of course).

BTW, looking at the actual track of ACA759, while using the 28R FMS bridge visual approach (as confirmed in ATC communication), they didn’t turn to exit waypoint F101D but went direct toward the runway area (their actual flight path was direct to Taxiway C??) from SAMML. Had they made that FMS turn and intersected F101D they would have been more directly aligned with 28R and the entire landing view would have been more clear (along with ND confirmation of 28R directly ahead). All just my opinion.

mwscan wrote:

I'm not a pilot, but I do notice things about communication.

Quote: Air Canada pilot: Tower Air Canada 759 I can see lights on the runway there. Can you confirm we’re clear to land?

This is somewhat ambiguous.

"Lights on the runway" can be interpreted as "runway lights".

or

"Lights on the runway" can be interpreted as simply "lights" which could be the landing lights and nav lights of the aircraft lined up on the taxiway.

The fact that he said, "Can you confirm we're clear to land" might indicate that he had some concern that things looked out of the ordinary. He could have/should have been more specific in what he meant by "Lights on the runway".

Seems no other folks were interested in addressing your question, or I have missed their answer, if you did receive one.

Anyway, I wish to inform you that usually if a pilot states he/she sees lights on the runway, that aren't referring to anything other than just that -- on the runway.

Not on the side, or over there, or anywhere but -- on the runway.

True, I've been grounded for a fair number of years, but some things one just doesn't forget and whenever I heard anyone discussing something on the runway, that was exactly what they meant -- on it.

I suspect you'd be surprised at some of the stories folks can tell you about an abort because something weird was on the active. Not everyone gets to fly into those nice, fancy airfields where all sorts of safety this and that is around to help you. Sometimes you just have to fend for yourself, because ole' Smokey himself might just be on the active having a smoke. Oh yeah, and Smokey knows darn well you'll be going around and will wait for him to finish.

atypical wrote:

I hate doing this on airliners because no one likes their exuberance dismissed, but lets look at what we have:

A doctored recording of the ATC conversation as the only evidence. A retired "aviation expert" saying this could have been worse than the Canary Islands. So this news report that is the primary source of information is clearly showing a sensationalist bias. Aside from that report there is no evidence indicating that anyone was in any danger at any time. In fact there is no evidence to indicate that PIC made an error of judgment worth including in his record. So this whole event is as yawn worthy as it is disaster averted. At least wait for there to be some blood in the water before people are condemned.

You cannot be serious. A hundred and some-odd feet above four other planes and descending and no one was in danger at any time? I would submit that this is the type of gung-ho thinking that allows glaringly dangerous situations to persist until people get killed.

I never get how this can possibly happen, both pilots need to be majorly distracted for this to occur. If a junior FO screws up like this, okay; but how can the captain possibly not notice this? Absolutely incomprehensible to me.

NBC must be in charge of training AC crews the difference between runways and taxiways...

https://youtu.be/_gIjTjW5dyM

ryanov wrote:

A different thread on this forum says:

"From Air Disaster Volume 3:

CFM56-5:
29% N1 to 67%: 4 seconds
67% N1 to 83%: 1 second"

And also that there is an FAR that specifies a maximum time:

"There is an industry standard. They must accel from flight idle to 95% rated thrust in 5 seconds or less. It is found in FAR Section 33.73(d)."

The series of books written by Macarthur Job were brilliant, made it easy to understand what happened. Do miss that guy.

CO953 wrote:

Exactly. We are saying the same thing in different languages. At Aspen they never confirmed acquisition of the runway. Read the transcript.... there was never an agreed affirmative. Yes, the charter customer in the cockpit made it worse. The controller saw the jet come out of the twilight haze pointed directly at the cliff and hit the crash button. The pilots committed to the landing and never between them agreed that they saw the runway.\

Thank God that the Air Canada crew had some visibility. Otherwise at SFO we have the worst crash in world history, potentially killing 1K people

Try this again at SFO with dense fog and you have five piles of ash on the taxiway.

If there was fog then it would have been an ILS approach. You don't see the runway at minimums then you go around.

This begs the question if you only are using 28R and not 28L at night why are they not doing ILS approaches at that time. Surely this can't create a huge amount of delays at SFO at that time and if it is that busy why aren't they using 28L for takeoffs. I know 28R is a longer runway but not by much and it should be able to accommodate 99% of long haul flights at the midnight hours

aviationaware wrote:

I never get how this can possibly happen, both pilots need to be majorly distracted for this to occur. If a junior FO screws up like this, okay; but how can the captain possibly not notice this? Absolutely incomprehensible to me.

I'm not a pilot but, having a relative who is now a retired pilot and having grown up in a time when cockpit visits were freely allowed I have had the privilege to be in the cockpit for landings (and take offs), all over the world, from LAX to JFK, to SIN, to AKL, to LHR, to LIS, you name it I've done it. Speaking strictly as an observer, from a fair distance out it's amazing that missed runway approaches don't occur more often. To a untrained observer like me it only becomes obvious that you are or are not on the right path, when you're pretty close. So at night, I can imagine how easy it might be to make such a mistake, and the fact that it doesn't happen more often is a testament to how well trained, pilots all over the world are.
Having said that, even if 99.999% of approaches are correct there's really no excuse in this day and age for that 0.001% failure, or whatever it really is. I cannot imagine that there is no cheap and reliable technology in place to alert pilots when they're lined up on a taxiway. I'm thinking that a proximity sensor on the edge of a taxiway, that would detect an approaching airplane and light up some lights should do the job.

So, what do those two pilots got in store for them? Any chance they will be able to pilot a commercial plane again, or is it "one strike and you're out" in the aviation business?

Doesn't the A320's automated ILS landing systems prevent this sort of thing from occurring? (unless the pilot switched them off)

I believe it was in the Michael Creighton book, Airframe, where a plane either lost control or lined up on a taxiway (I believe the former) and the pilot of the plane holding short saw it coming and tried to spool up his engines and move the plane but it took too long and the landing aircraft struck the holding aircraft. I can't imagine how helpless it would feel to be the UA1 guy watching this unfold.