SFO Incident

[Seeker]

6 hours ago, Tango Niner said:

I firmly believe that even in the total absence of warnings given by ATC or anyone else, this AC crew would have eventually initiated a last minute (and, yes, rather hair-raising) goaround, rather than settling their bird into a lineup of waiting aircraft.

I agree. 

Let me tell a personal story;  one time I was flying into ORD and from several miles back we could see a situation developing on our intended runway - one guy slow to clear and another slow to go to position.  We didn't go around at the first indication that it wasn't going to work, we continued down the slope.  PF and PM discussed the probability of a G/A but continued the approach until it really, really wasn't possible to land and then we did our G/A - about the same time as the TWR called for it.  An outside observer might comment about how "close" we came to disaster or about how we "didn't see" the aircraft sitting on the runway in front of us, while in reality we saw it from miles back.  Sometimes situations resolve themselves, sometimes they don't.  I've had YYZ TWR give me landing clearance below minimums - several times actually.

I can't say for certain what happened in SFO but remember the very first thing that happened was the query by the approaching aircraft about lights on the runway so obviously they were aware that something wasn't right.  I think that they were aware something didn't look right but because of the lighting conditions, approach offset, crosswind or some combination of factors couldn't quite reconcile the information so they queried TWR and continued.  TWR said, "everything is fine."  So now we're supposed to believe that they would just turn off their brains and land on top of four wide-bodies without looking out the window again or any further concerns?  Well, maybe, but more likely is that they continued on their approach being ready for either a G/A or landing depending on how the situation resolved itself.

These guys may have been completely in the wrong, approaching a taxiway instead of a runway, but the fact that the G/A was done from a low altitude, alone, doesn't prove that they weren't aware that something wasn't right.

[mo32a]

The offset approach is corrected at 4 miles final, they were much closer than that. In fact with 28L out of service and a visual approach there was no need for them to be offset at all. I would think even if you are doing a visual you would tune the ILS up and see immediately that you weren't where you were supposed to be.

[Vsplat]

Given the time of arrival, they were likely cleared for a noise abatement visual.  The expectation would be that they would follow the cleared approach routing until cleared otherwise. 

SFO is not always pristine in their clearances, perhaps there was some confusion here. Certainly the way in which tower replied when the crew queried traffic was less clear than it could have been  

Vs

[gator]

I'm not convinced that all of the information supposedly supplied by authorities is correct. For example, they say that the flight overflew the first two aircraft at 100 ft, the 3rd by 200 ft and the 4th by 300 ft. 

With 100 ft between aircraft and a 200 ft body length, the 4th aircraft is only 400 ft from the second aircraft.

So the aircraft in the go around climbed 200 ft in 500 ft (the distance between the 2nd and 4th aircraft with 100 ft between them). That's about 5,000 ft/min (23 degree angle of climb... meaning more than 20 degree body angle with flaps out) at 125 kts. The A320 is pretty dramatic in go around mode, but I don't think it did 5,000 ft/min.

Bottom line... there's a lot of holes in the (sparse, at best) available information. You know the drill....

 

[DEFCON]

"Bottom line... there's a lot of holes in the (sparse, at best) available information."

 

Couldn't agree more!

 

[moeman]

https://ntsb.gov/investigations/Pages/DCA17IA148.aspx

59 feet above ground level...yikes.

Quote

The incident pilots advanced the thrust levers when the airplane was about 85 ft agl. FDR data indicate that the airplane was over the taxiway at this time, approaching the vicinity of taxiway W.

o    At 2356:04 PDT, ACA759 reappeared on the local controller’s ASDE-X/ASSC display as it passed over the first airplane positioned on taxiway C.

o    About 2.5 seconds after advancing the thrust levers, the minimum altitude recorded on the FDR was 59 ft agl.

 

[Guest]

2 hours ago, moeman said:

https://ntsb.gov/investigations/Pages/DCA17IA148.aspx

59 feet above ground level...yikes.

 

the most important observation was this and it does appear that the crew had taken corrective action well before being advised to "go around":

o    The incident pilots advanced the thrust levers when the airplane was about 85 ft agl. FDR data indicate that the airplane was over the taxiway at this time, approaching the vicinity of taxiway W.

o    At 2356:04 PDT, ACA759 reappeared on the local controller’s ASDE-X/ASSC display as it passed over the first airplane positioned on taxiway C.

o    About 2.5 seconds after advancing the thrust levers, the minimum altitude recorded on the FDR was 59 ft agl.

o    At 2356:10 PDT, the local controller directed ACA759 to go around. The airplane had already begun to climb at this point (see figure 4).

[UpperDeck]

2 hours ago, moeman said:

https://ntsb.gov/investigations/Pages/DCA17IA148.aspx

59 feet above ground level...yikes.

 

Thanks, Moeman. This info is very similar to that published earlier by another source. It was the second aircraft in line...not UA01....that was closest to impact as altitude deteriorated prior to climb. Those waiting aircraft were on the taxiway and the AC flight passed DIRECTLY overhead

[UpperDeck]

18 minutes ago, Malcolm said:

the most important observation was this and it does appear that the crew had taken corrective action well before being advised to "go around":

o    The incident pilots advanced the thrust levers when the airplane was about 85 ft agl. FDR data indicate that the airplane was over the taxiway at this time, approaching the vicinity of taxiway W.

o    At 2356:04 PDT, ACA759 reappeared on the local controller’s ASDE-X/ASSC display as it passed over the first airplane positioned on taxiway C.

o    About 2.5 seconds after advancing the thrust levers, the minimum altitude recorded on the FDR was 59 ft agl.

o    At 2356:10 PDT, the local controller directed ACA759 to go around. The airplane had already begun to climb at this point (see figure 4).

Malcolm...

From my perspective, what you have quoted indicates that the controller was "late" in responding to the situation though the aircraft had disappeared from his display and his response was almost immediate once it reappeared. The " go round" may have been initiated prior to direction by the controller but at a flight level of 85 feet OVER THE TAXIWAY!!

[gator]

 

So, is is radar altitude or barometric altitude? Or GPS altitude?

If it is radar altitude it would be reading the distance to the closest hard object. If it was over an aircraft on the ground (as stated), it would be reflecting off the closest part of that aircraft, possibly the tail, so the 59 feet indicated on the RA would translate into 89+ feet AGL if reflecting off the body and maybe as much as 120 feet AGL if reflecting off the tail. 

If it was barometric altitude, nobody has any right to start quoting to the foot.

If it was GPS altitude (assuming this aircraft was equipped... not sure),  the error there can be in the 10s of feet, too.

edit.... not to mention that if they were at 59 feet directly over a 777, they would have impacted the tail.

[blues deville]

4 hours ago, gator said:

 

So, is is radar altitude or barometric altitude? Or GPS altitude?

If it is radar altitude it would be reading the distance to the closest hard object. If it was over an aircraft on the ground (as stated), it would be reflecting off the closest part of that aircraft, possibly the tail, so the 59 feet indicated on the RA would translate into 89+ feet AGL if reflecting off the body and maybe as much as 120 feet AGL if reflecting off the tail. 

If it was barometric altitude, nobody has any right to start quoting to the foot.

If it was GPS altitude (assuming this aircraft was equipped... not sure),  the error there can be in the 10s of feet, too.

edit.... not to mention that if they were at 59 feet directly over a 777, they would have impacted the tail.

You're confusing the jury.....and I'd want you on my defence team.

From what I read in this short NTSB report all altitudes units of measurement were clearly stated. 

[UpperDeck]

9 hours ago, gator said:

 

So, is is radar altitude or barometric altitude? Or GPS altitude?

If it is radar altitude it would be reading the distance to the closest hard object. If it was over an aircraft on the ground (as stated), it would be reflecting off the closest part of that aircraft, possibly the tail, so the 59 feet indicated on the RA would translate into 89+ feet AGL if reflecting off the body and maybe as much as 120 feet AGL if reflecting off the tail. 

If it was barometric altitude, nobody has any right to start quoting to the foot.

If it was GPS altitude (assuming this aircraft was equipped... not sure),  the error there can be in the 10s of feet, too.

edit.... not to mention that if they were at 59 feet directly over a 777, they would have impacted the tail.

Question.....the report states that the data was extracted from the FDR. Is height above ground being recorded measured by radar, barometer AND gps ( assuming the aircraft is so equipped)? Clearly, it would be wrong of the NTSB to report a measurement extracted from only one source if two others were available and differed.

[gator]

I think that the NTSB/FAA have been more sloppy n this incident than I have ever seen before.

They refer to one altitude as about 85 feet AGL (no reference to FDR or the data source itself (baro, RA, GPS),  then refer to the minimum altitude recorded on the FDR (again, no reference to the data source) as 59 feet (to the foot... not "about"). If the 85 feet was off the FDR and sourced from the RA, it shouldn't have been "about" and if it wasn't, it's from radar (5 second update) or ADS-B... which would have been more timely than radar, but also recorded on the FDR. Not sure the vertical accuracy reported by ADS-B but, I believe it uses barometric altitude. 

They let the idea that the crew waited until the tower told them to go around before they initiated the missed approach sit out there for the better part of a month.  Now, buried in an interim report is the fact that the tower was a couple of seconds behind the crew in this, but despite having things timed to the second, they conveniently omit the time at which the crew initiated the go around and uses the phrase "about 2.5 seconds after advancing the thrust levers". 

UD, it doesn't state that the data was extracted from the FDR. It only refers to specific items that were from the FDR. The source of other items... 85 feet, 2.5 seconds is not stated.

I'm not defending the pilots completely, but there is a lot more to this than just inaccurate data. Fatigue, approach design, the fact that there was only one controller in the tower, physio-psychological sensations caused by go-around acceleration... the latter played a role in Rostov and a very close call in France. I wonder if the hero-izing of the controller was allowed to continue in an effort to relieve pressure off the FAA for reducing tower manpower?

 

[Guest]

Latest from Global Link contains some video.  http://globalnews.ca/news/3644687/air-canada-near-miss-san-francisco/

August 3, 2017 3:33 am

Images show how dangerously close Air Canada jet came to crashing in San Francisco

By David Koenig The Associated Press

An Air Canada flight carrying 140 passengers from Toronto to San Francisco nearly had a dramatic end when the aircraft came close to landing on a crowded taxiway.

 

Newly released data and photos show how shockingly low an Air Canada jet was when it pulled up to avoid crashing into planes waiting on a San Francisco International Airport taxiway last month.

The Air Canada pilots mistook the taxiway for the runway next to it and flew their jet to just 59 feet (18 metres) above ground before pulling up to attempt another landing, according to National Transportation Safety Board information released Wednesday.

That’s barely taller than the four planes that were on the taxiway when the incident occurred late at night on July 7.

 

<img class="story-img" src="https://shawglobalnews.files.wordpress.com/2017/08/la308-82_2017_201750_high1.jpg?quality=70&#038;strip=all&#038;w=512&#038;h=288&#038;crop=1" alt="This composite of images released by the National Transportation Safety Board (NTSB) shows Air Canada flight 759 (ACA 759) attempting to land at the San Francisco International Airport in San Francisco on July 7, bottom. At top is a map of the runway created from Harris Symphony OpsVue radar track data analysis. At center is from a transmission to air traffic control from a United Airlines airplane on the taxiway. The bottom image was taken from San Francisco International Airport video and annotated by source.">

This composite of images released by the National Transportation Safety Board (NTSB) shows Air Canada flight 759 (ACA 759) attempting to land at the San Francisco International Airport in San Francisco on July 7, bottom. At top is a map of the runway created from Harris Symphony OpsVue radar track data analysis. At center is from a transmission to air traffic control from a United Airlines airplane on the taxiway. The bottom image was taken from San Francisco International Airport video and annotated by source.

Pilots in a United Airlines plane alerted air traffic controllers about the off-course jet, while the crew of a Philippine Airlines jet behind it switched on their plane’s landing lights in an apparent last-ditch danger signal to Air Canada.

NTSB investigators said they have not determined probable cause for the incident that came within a few feet of becoming one of the worst disasters in aviation history.

“It was close, much too close,” said John Cox, a safety consultant and retired airline pilot.

The investigators said that as the Air Canada jet approached the taxiway just before midnight after a flight from Toronto, it was so far off course that it did not appear on a radar system used to prevent runway collisions.

VIDEO: Air Canada flight involved in close call with other planes

Those systems were not designed to spot planes that are lined up to land on a taxiway – a rare occurrence, especially for airline pilots. But the Federal Aviation Administration is working on modifications so they can, agency spokesman Ian Gregor said.

Both pilots of the Air Canada Airbus A320 jet were very experienced. The captain, who was flying the plane, had more than 20,000 hours of flying time, and the co-pilot had about 10,000 hours.

The pilots told investigators “that they did not recall seeing aircraft on taxiway but that something did not look right to them,” the NTSB said.

Investigators could not hear what the Air Canada captain and co-pilot said to each other during the aborted landing because their conversation was recorded over when the plane made other flights, starting with a San Francisco-to-Montreal trip the next morning. Recorders are required to capture only the last two hours of a plane’s flying time.

Peter Fitzpatrick, a spokesman for Air Canada, declined to comment, citing the ongoing investigation.

VIDEO: Air Canada has close call in San Francisco

© 2017 The Canadian Press

New images show Air Canada jet was closer than believed to near disaster in San Francisco

'It was close, much too close,' says aviation safety consultant

The Associated PressPosted: Aug 03, 2017 2:47 AM ET Last Updated: Aug 03, 2017 8:19 AM ET

The National Transportation Safety Board has released photos taken from San Francisco International Airport video showing the July 7 "near miss" when Air Canada Flight 759 came close to landing on a taxiway with four airplanes on it instead of on the runway. (National Transportation Safety Board )

201 shares

Facebook

Twitter

Reddit

Google

Share

Email

Related Stories

• Air Canada jet was 30 metres from colliding with 2 planes in San Francisco: TSB report
• 'Where's this guy going?' Air Canada plane narrowly avoids disaster in San Francisco

External Links

• NTSB report on Air Canada Flight 759

(Note: CBC does not endorse and is not responsible for the content of external links.)

Newly released data and photos show how shockingly low an Air Canada jet was when it pulled up to avoid crashing into planes waiting on a San Francisco International Airport taxiway last month.

The Air Canada pilots mistook the taxiway for the runway next to it and flew their jet to just 18 metres above ground before pulling up to attempt another landing, according to National Transportation Safety Board information released Wednesday.

That's barely taller than the four planes that were on the taxiway when the incident occurred late at night on July 7.

• Air Canada plane narrowly avoids disaster in San Francisco

Pilots in a United Airlines plane alerted air traffic controllers about the off-course jet, while the crew of a Philippine Airlines jet behind it switched on their plane's landing lights in an apparent last-ditch danger signal to Air Canada.

Air Canada Flight 759 approaches the taxiway. (National Transportation Safety Board )

Air Canada Flight 759 flies over the first airplane waiting on the taxiway. (National Transportation Safety Board )

Air Canada Flight 759 had begun to climb as air traffic control in San Francisco told it to 'go around.' (National Transportation Safety Board )

[Johnboy]

No where near being a subject matter expert but offering this  in the interest of discussion only. The NTSB reports AC759 was cleared for the FMS bridge visual runway 28R approach. This approach places the aircraft in a right of centerline (28R) on a somewhat converging course to the runway (see attached plates) vs an ILS 28R approach which of course places you on the LOC/GS. The weather conditions were VMC and SFO was in a somewhat unusual configuration with DEP and ARR on 28R. I assume that the runway lighting system for 28L was off (reduced setting) due to the runway closure. All of these factors sets the aircraft up in a position to line up with taxiway C bearing in mind close proximity of the parallels at SFO 28/01's. The crew is convinced there are lined up with the correct runway (visual approach) with no ILS backup and the aircraft on the taxiway are more than likely invisible to the crew until closer in (aircraft taxi lights would be off). Please note AC would be using Jeppesen charts but the comparison would be a very close comparison.

Not sure at what point back would the blue lighting of the taxiway become evident to the PF. The visual and the (jedi mind trick) leading you into thinking all is good. This video of the FMS bridge visual runway 28R approach is DVFR fast forward to the 8:54 mark.

 

 

 

 

[DEFCON]

Does it really matter if it was 3', or fifty, the landing aircraft came very close to others?

To my eye the airport design / layout is very poor. When the push is on to keep aerodrome performance up the shortcomings in its design seem to create a number of traps that can catch an unsuspecting crew. In this case ATC's use of a parallel taxiway that is positioned far too close to the active runway helps to set the stage for disaster, a situation that's likely exasperated by darkness. Although it's not at issue here, inclement weather would make an already challenging landing environment even more difficult.

The 'media report claims the investigation has revealed a twelve mile radar blind spot? I find it difficult to believe a hole in coverage that large could go undiscovered until this incident occurred.

I'm still trying to understand the apparent absence. or non-use  of ILS guidance?

Arguably, the biggest question that needs to be answered centers around the optics that confused the AC crew and any factor, fatigue for instance, that influenced, or may have aggravated  the situation.

  

 

 

 

[Guest]

What they should have seen.....  28R at night ...... 

[DEFCON]

Thanks Malcolm, but this photo doesn't present an entirely accurate representation of the lighting configuration, nor the perspective view of the runway environment the pilots would have had during the approach on the night in question, but it's close enough to make one wonder what influence may have led an obviously experienced crew so far off the appropriate track.

I wonder if earlier incident reports from other pilots exist that may shed some light on this incident?

 

   

[Guest]

9 minutes ago, DEFCON said:

Thanks Malcolm, but this photo doesn't present an entirely accurate representation of the lighting configuration, nor the perspective view of the runway environment the pilots would have had during the approach on the night in question, but it's close enough to make one wonder what influence may have led an obviously experienced crew so far off the appropriate track.

I wonder if earlier incident reports from other pilots exist that may shed some light on this incident?

 

   

Is this closer? This one appears to be from the cockpit.

[Johnboy]

2 hours ago, DEFCON said:

I'm still trying to understand the apparent absence. or non-use  of ILS guidance?

 

  

 

 

 

If I am reading the approach plate correctly the only guidance available is the GP for 28R and no LOC available for guidance. But not privy to AC SOP for flying this type of approach. Flying the LOC would put left of the desired the desired flight patch (Quiet App).

[DEFCON]

"Is this closer?"

Short story ... no.

I'm certain the crew didn't intentionally cause the incident, which causes me to focus on the influence of wind on the track of the aircraft during the offset portion of the approach and the 'changing' visual perspective of the pilots; the crew query regarding the landing environment is telling us something important.

A while back UD pointed out the fact that thousands of pilots have flown that approach in similar circumstances without incident. Accordingly, fatigue stands out as perhaps the most compelling causal factor. Unfortunately, we'll have to wait for the report.

 

 

[boestar]

am i missing something?  The runway CLEARLY has approach lighting and the taxiway does not.  How does one miss that little tidbit of visual cues?

 

[DEFCON]

Thanks Johnboy

Maybe I'm confused, but If the visual transition to the final approach path occurs at a prescribed point / distance from the 28R threshold, as depicted in Malcolm's second photo, wouldn't the 28R ILS be usable to ensure the aircraft is on the appropriate flight path?

 

 

 

[DEFCON]

The answer to that question remains something of a mystery Boestar.

 

[blues deville]

On 2017-08-03 at 2:48 PM, Johnboy said:

If I am reading the approach plate correctly the only guidance available is the GP for 28R and no LOC available for guidance. But not privy to AC SOP for flying this type of approach. Flying the LOC would put left of the desired the desired flight patch (Quiet App).

There is a slight difference in the depiction of this approach between your FAA version and Jeppesen. While the FAA clearly shows the location of the 28L/R localizers, it doesn't show the transistion (text only info) to it from the SFO 095R, which should occur at 6 DME or passing the San Mateo Bridge. If somehow you got busy looking out and forgot to make the transistion to the LOC, you would end up pointing at the Charlie taxiway and not the runway center lines. 

t