Boeing 777 plane crash-lands at San Francisco airport

[EyesWideOpen]

The PAPI was working, which is just as accurate as a ILS glide slope, but neither would have been needed for this approach anyway.

Like AF 447 this accident appears to be another shocking example of airline pilots forgetting (or never having really learned) the basics when something goes wrong with the automation.

Massive difference between this accident and the Air France one. So big a difference I find it hard to draw a comparison with them as you have.

Flight was on final approach in clear weather operating under VFR. One of the fundamental basics of final approach under VMC is the positioning of the approach point through the cockpit windows, and adjusting your speed and attitude accordingly.

I disagree . When the details have come out basically spelling out pilot error in this crash , the Air France flight was the very first thought in my mind. Both planes had flight crews lacking in basic flying skills. The position the plane was actually in is irrelevant, whether on final approach or on cruise like the Air France flight. Air France came out looking very very bad, and later they did a massive revamp of their training program. With the Air France flight, every body on board paid for the incompetence of the crew with their life. The fact that nearly everybody survived this crash is truly astonishing....

Edited July 8, 2013 by EyesWideOpen

[Mosha]

NTSB Initial report. The pilot was familiar with SFO

http://avherald.com/h?article=464ef64f&opt=0

Edited July 8, 2013 by Mosha

[TallGuyJohninBKK]

Just not with landing 777s there....

On Jul 8th South Korea's Ministry of Transport reported the captain (43, ATPL, 9,793 hours total) of the ill-fated flight was still under supervision doing his first landing into San Francisco on a Boeing 777, although he had 29 landings into San Francisco on other aircraft types before. He was supervised by a training captain with 3,220 hours on the Boeing 777, all responsibilities are with the training captain.

Edited July 8, 2013 by TallGuyJohninBKK

[skippybangkok]

Here is a graph, the far left of the "U" curve is the "back end of the power curve"

Very un forgiving and just a few knots lower than you stall speed, you can apply full thrust, and it will do diddly squat.

As far as I know the only way out is nose down and full thrust, which is not an option 100 ft above the Tarmac

[EyesWideOpen]

It is interesting to compare this crash with the only other 777 crash.

Due to icing, the engines on that plane would not spool back up during the final descent, with the plane looking like it would come in well short of the runway , and certainly killing almost everybody on board. The British Airways captain , within a few seconds ,made the brilliant decision of raising the flaps. The plane dropped in altitude due to less lift, but went further due to less drag. The plane just made it to a landing on the dirt next to the landing strip, and every single person walked away alive. This was a classic example of piloting skill, versus guys trained to turn the autopilot on and off....

[dickyknee]

There was some speculation in another thread of that the 2 passengers that died, one may have been killed by a firetruck.

now all speculation aside, no blame should be placed on the emergency services in response to any emergency.

[Maestro]

...meanwhile here is the ATC, probably a little hard to follow unless your one our TV Experts:

http://wandr.me/Audio/AAR214-KSFO-Crash.mp3

Do I understand correctly that there was talk of an emergency already when the plane was at a height of 3,500 to 3,000 metres? "Emergency vehicles are responding". "Emergency vehicles are on their way"

Huh. What are you reading. 7 second prior to impact, they mentioned need more speed. 4 second prior to impact stick shaker. 1.5 seconds prior to impact, said going around.

As indicated, there was some delay in power. Should have been at full throttle at 7 seconds since I believe the recorded said something like 123 knots. Stick shaker would definitely be full power. I don't think spool up would have been 3 seconds before gaining some speed if all went accordingly, but I am not sure and he pulled nose up which would further bleed speed probably trying to stretch the distance anyway he could at very last moments before impact.

Is that a recording of the communication between the tower and more than one aircraft on landing approach? It seems so but I did not realise this earlier. It is apparently the pilot of another plane that gave those hight levels of 3500 and 3100 (starting at about 42 seconds into the recording). Sorry about the mix-up.

[Dancealot]

There was some speculation in another thread of that the 2 passengers that died, one may have been killed by a firetruck.

now all speculation aside, no blame should be placed on the emergency services in response to any emergency.

Of course, no suggestion was made to blame. I just wanted to raise awareness.

And apparently, I have.. Thank you, dickyknee.

[Dancealot]

The investigation about what really happened continues. Rumours are a fire engine accidentally ran over one of the survivors...

http://www.nbcdfw.com/news/national-international/Fire-Truck-May-Have-Run-Over-Asiana-Plane-Crash-Victim-214546021.html

[dickyknee]

There was some speculation in another thread of that the 2 passengers that died, one may have been killed by a firetruck.

now all speculation aside, no blame should be placed on the emergency services in response to any emergency.

Of course, no suggestion was made to blame. I just wanted to raise awareness.

And apparently, I have.. Thank you, dickyknee.

I wasn't going to reply to an off-topic post in another thread (now deleted) so I replied where it was on-topic.

and to quote the article posted "Fire Truck May Have Run Over Asiana Plane Crash Victim"

no suggestion of blame, of course.

[Dancealot]

I wasn't going to reply to an off-topic post in another thread (now deleted) so I replied where it was on-topic.

and to quote the article posted "Fire Truck May Have Run Over Asiana Plane Crash Victim"

no suggestion of blame, of course.

Good for you.

I still wonder what really happened...

[wooloomooloo]

I still wonder what really happened...

A bad accident happened and incidental events are unfortunate.

On this occasion, I can only wonder that the majority survived. Another day it could have been 100% fatality. That's of course no benefit to the two confirmed dead girls.

[ricklev]

Why don't they print what the flight crew is saying? A million news stories and not a one mentions that the the crew is alive and fully capable of explaining their experience.

Edited July 8, 2013 by ricklev

[Mrjlh]

Just got back from the NTSB hour long briefing. Yes I Live here in SF and was allowed in. A ton of information to sort through but very precise.

1. They are a little surprised at the current news reports that are going around that are far from the facts. Because not all the facts are in. Made a comment to not report speculation.

2. Today they are interviewing the pilots and should have a "preliminary" report for tomorrow's briefing.

3. Comments about aborting the landing already reported in the news are from the "flight deck voice recorder" are not "official" requests to the tower. These where only discussions among the crew to abort the landing. These conversations are private and privileged. They were reported yesterday by the NTSB only to help clarify some of the actions by the crew. Radio transitions to the tower are public and open. Two totally different recording devices.

4. The Flight recorders were taken back to Washington DC and run on a "simulator", ONLY to gain a "PRELIMINARY" idea of the events. There are 1400 different data points that need to be graphed, plotted and analyzed. Only from putting all the evidence together will they have an accurate and complete picture of what happened. They did give out time, altitude, and speeds of the A/C on final approach which appeared to be normal until the last 30 seconds when they were obviously too slow and too low but in the last 7 seconds all indicators showed they were attempting to correct the problem. A lot happens in 30 seconds with a very heavy plane.

5. Because they were landing many alarms were turned off. Mainly the "close to terrain alarm" because they are landing. But it is to early to speculate until all the info is in.

6. They have 20 investigators working the crash.

7. She explained very much in detail how the crash occurred. How the tail hit first, brok off and left a big part of it in the water. How the plane broke up down the runway. Where all the main pieces ended up. How passengers got out. When and how the fire started. Pretty basic stuff but with a lot more detail too.

8. Tomorrow they will have another news conference. Just too much to absorb in one posting.

9. Just to show you how "rumors and speculation" get started, one reporter asked was it true the pilot had a heart attack before landing? NTSB spokeswoman handled it pretty good. "First I heard. Not true. Next question?

So every day we will get a little more truth about what happened. The press conference with the "First Responders" was very good too. That's another discussion.

So ThaiVisa rant, speculate, make facts up that have nothing to do with this accident. In others words... be yourselves in true fashion. Maybe I;ll report more real time...if I go again.

[khaosai]

Hi,

Seems the answer might be with the auto throttle flight mode annunciation. There are some traps for the unwary with the B777 system that may prevent the auto throttle wake up function from engaging in certain situations. With the auto throttle engaged in FLCH or when below 100ft on approach you do not get stall protection. If descending in VNAV SPD and HOLD annunciates you will not get stall protection.

They may well have got a sink rate caution if excessive descent rate was sensed and an airspeed low caution if below minimum manoeuvre speed.

The recommended technique on the B777 is to leave the auto throttle engaged during approach and landing, even whilst manually flying the aircraft. The technique on the B747 is slightly different where you would disconnect the auto throttle whilst manually flying the aircraft.

With regards to the communication being possible on the ground with no engines or APU, the aircraft battery will provide power to the emergency lighting and the Captains control display unit which contains the aircraft radios.

Line training with full fare paying passengers onboard is a common feature in airline flying. It's taken very seriously, with the actual training taking a very distant third behind the safety and commercial aspects of the job.

Edited July 8, 2013 by khaosai

[TallGuyJohninBKK]

Line training with full fare paying passengers onboard is a common feature in airline flying. It's taken very seriously, with the actual training taking a very distant third behind the safety and commercial aspects of the job.

Would that typically include having a pilot with only 40+ hours in 777s total be sitting in the captain's chair and controlling the landing with a fully laden plane at an airport where he'd never flown a 777 before?

Does that put the training "a very distance third behind the safety and commercial aspects of the job"?

I don't know.. I'm just asking... as an airline passenger.

[TallGuyJohninBKK]

Here's the NTSB video on YouTube of their latest Monday media briefing:

[NeverSure]

Line training with full fare paying passengers onboard is a common feature in airline flying. It's taken very seriously, with the actual training taking a very distant third behind the safety and commercial aspects of the job.

Would that typically include having a pilot with only 40+ hours in 777s total be sitting in the captain's chair and controlling the landing with a fully laden plane at an airport where he'd never flown a 777 before?

Does that put the training "a very distance third behind the safety and commercial aspects of the job"?

I don't know.. I'm just asking... as an airline passenger.

Yes, very standard. There isn't really a captain's chair. The plane can be fully flown from either seat. If the pilot in the right seat was the trainer, and the one in the left was the trainee, then the pilot in the right seat was the pilot in command and had full access to all instruments and controls.

40 hours is actually quite a bit of time. The guy had thousands of hours in heavies, so it wasn't as if he was a newbie. Here you have two highly experienced pilots and the SHTF. Either one could have firewalled the throttles 30 seconds out and it would all have been a non-issue. That's assuming there was no mechanical failure.

[TallGuyJohninBKK]

Some of Hersman's comments from the Monday media briefing:

--Re the interviews of the four pilots on the aircraft, Hersman said they wanted to have representatives from Asiana and the Korean transport agency onsite before NTSB proceeded with them Monday U.S. time.

--Hersman was asked but couldn't answer how many of the 4 pilots on the aircraft were in the cockpit at the time of the crash. Earlier, she also seemed to avoid specifying which of the two principal pilots that have been discussed was actually in control of the aircraft at the time of the crash.

--Hersman declined to answer questions about what the rules and standards are for things such as check pilot qualifications, and whether they vary from airline to airline, or not -- saying they wanted to complete the pilot interviews before commenting further.

--Hersman said although the NTSB cannot legally release the cockpit voice recordings of the conversations among the pilots, she said they would eventually release a written transcript of those recordings.

-------------------------------------------

--NTSB will be reviewing prior aircraft flights into SFO to determine any impacts from the airport's ILS outage or the runway construction project that was ongoing at SFO.

------------------------------------------

--Based on flight data recorder "early reads," she said the autopilot was disengaged at about 1600 feet or 82 seconds prior to impact. At about 500 feet of altitude and 34 seconds prior to impact, the airspeed was approx. 134 knots. At approx 200 feet and 16 seconds prior to impact, the airspeed was approx. 118 knots. At about 125 feet and 8 seconds prior to impact, the throttles started moving forward and airspeed was approx. 112 knots.

--About 3 seconds prior to impact, the flight data recorder recorded its lowest speed of 103 knots. At this time, the engines were at about 50% power and power was increasing. At impact, airspeed was approx. 106 knots. Hersman noted that the 777's designated speed for the point of cross the threshhold of the runway is supposed to be 137 knots, and that the Asiana craft was "significantly slower."

--Re news reports of a 4000 ft per minute descent rate for the crashed aircraft, she said, "There's no abnormally steep descent curve that's been detected in the data" they've reviewed thus far.

-------------------------------------------

--"We have heard that there were some problems inside of the aircraft... There have been some interviews with flight attendants and witnesses that slides deployed inside of the aircraft. We need to understand why that happened."

-------------------------------------------

--"There is also debris from the seawall located several hundred feet up the runway." As well as a lot of debris from the tail of the plane in the water near the seawall.

--Tail/rear area of the aircraft where the two fatalies were seated: "It's an area where we're seeing a lot of the critical or serious injuries that occurred as well."

Edited July 9, 2013 by TallGuyJohninBKK

[TallGuyJohninBKK]

Either one [pilot] could have firewalled the throttles 30 seconds out and it would all have been a non-issue. That's assuming there was no mechanical failure.

According to Hersman's account I posted above, the plane was already below designed landing speed and on a slowing trajectory 34 seconds and 500 feet of altitude before impact.

So I guess the couple of hundreds of million dollar question is: Why didn't they -- prior to the 8 seconds before impact and 125 feet of altitude when throttles began moving forward?

[MaxYakov]

Either one [pilot] could have firewalled the throttles 30 seconds out and it would all have been a non-issue. That's assuming there was no mechanical failure.

According to Hersman's account I posted above, the plane was already below designed landing speed and on a slowing trajectory 34 seconds and 500 feet of altitude before impact.

So I guess the couple of hundreds of million dollar question is: Why didn't they -- prior to the 8 seconds before impact and 125 feet of altitude when throttles began moving forward?

There are the requirements of a stabilized approach by the Flight Safety Foundation (see link). There are those of the opinion that it was not achieved by the aircraft during its approach. From what I understand, absence of a stabilized approach might have required a Go-around.

Ref: http://www.boeing.com/commercial/aeromagazine/aero_18/touchdowns_tab1.html

Here's a pertinent entry in the Wiki 'Go-around' page: http://en.wikipedia.org/wiki/Go-around

"Many airlines and aircraft operators state a list of conditions that must be satisfied so that a safe landing can be carried out. If one or more of these conditions cannot be satisfied then a go-around should be considered in some cases and must be carried out in others. This list is usually written in the operations manual which has to be approved by the relevant aviation authority (CAA in the UK, FAA in the United States). The operator's list of conditions is not exhaustive; pilots use their individual judgment outside of this scope."

There is a lot of speculation about why there may have been resistance to aborting the approach earlier by the pilot flying or on recommendation by another. The pprune[dot]org (Professional Pilot's Rumor Network) is up to 52 pages of discussion of this incident.

[Asiantravel]

I have heard reports that there were as many as four pilots on board this flight.

But whether that is correct or not, ultimately, the main question for me is what was the check pilot doing? Surely it was his job to make sure this kind of situation didn't arise?

[SeaVisionBurma]

The PAPI was working, which is just as accurate as a ILS glide slope, but neither would have been needed for this approach anyway.

Like AF 447 this accident appears to be another shocking example of airline pilots forgetting (or never having really learned) the basics when something goes wrong with the automation.

Massive difference between this accident and the Air France one. So big a difference I find it hard to draw a comparison with them as you have.

Flight was on final approach in clear weather operating under VFR. One of the fundamental basics of final approach under VMC is the positioning of the approach point through the cockpit windows, and adjusting your speed and attitude accordingly.

I disagree . When the details have come out basically spelling out pilot error in this crash , the Air France flight was the very first thought in my mind. Both planes had flight crews lacking in basic flying skills. The position the plane was actually in is irrelevant, whether on final approach or on cruise like the Air France flight. Air France came out looking very very bad, and later they did a massive revamp of their training program. With the Air France flight, every body on board paid for the incompetence of the crew with their life. The fact that nearly everybody survived this crash is truly astonishing....

These accidents are so vastly different I still find it impossible to compare them.

A330-200 series vs Boeing 777.

One in the middle of the night, in pitch black conditions over the middle of the Atlantic Ocean at high altitude - the other in clear daylight in the middle of the day at ground level.

One involved the crew having obvious spatial disorientation and giving incorrect control inputs to match the actual state of flight - the other was a controlled flight into terrain.

The earlier post I responded to says they are comparing it to Air France 447 because it is another case of the pilots forgetting the basics when something goes wrong with the automation. I don't accept that premise for the 777 accident, which happened in clear daylight on final approach under full control.

Just my two cents worth.

[negreanu]

The earlier post I responded to says they are comparing it to Air France 447 because it is another case of the pilots forgetting the basics when something goes wrong with the automation. I don't accept that premise for the 777 accident, which happened in clear daylight on final approach under full control.

Really? Are you aware of which Flight modes (Automation) were engaged or not engaged at the time of the incident. Was the automation at that point set incorrectly for the situation - AThrust mode ? FLCH mode?

Automation may yet play a role in this incident. Don't forget automation failure is different to automation mis-selection or mis-understanding of the system all could lead to a incident.

Clear day does not mean no automation modes are engaged. Ref Vertical profile management and Thrust control and lateral guidance.

The information will be released in due course.

Edited July 9, 2013 by negreanu

[hanno]

How strange re the deaths of the teenagers and their seat locations.

I'm no flying expert... But I've always heard that supposedly sitting the rear of an aircraft is supposed to be one of the safer or safest locations.

I guess it's not so safe if the plane crash lands tail first and the tail of the plane breaks off...

Apparently at least one, if not both girls, were already out of the plane and got hit by an emergency vehicle....

[skippybangkok]

Automation or not .... Beside the point.

I play on the FBT 737 simulator on Silom road regularly - from my first time, it was clear , when u select flaps 30, the speed drops off very quickly. Based on Vref 142, the worst I hit was about 125 knots. I know now. Select flaps 30, need to ( manual landing ) throttle up to 70%, for about 10 s, and then pull back to about 55% to not drop below about 140 knots.

Maybe he thought he was in Auto throttle and forgot a basic scan ie speed, height, vs and attitude

[MaxYakov]

The PAPI was working, which is just as accurate as a ILS glide slope, but neither would have been needed for this approach anyway.

Like AF 447 this accident appears to be another shocking example of airline pilots forgetting (or never having really learned) the basics when something goes wrong with the automation.

Massive difference between this accident and the Air France one. So big a difference I find it hard to draw a comparison with them as you have.

Flight was on final approach in clear weather operating under VFR. One of the fundamental basics of final approach under VMC is the positioning of the approach point through the cockpit windows, and adjusting your speed and attitude accordingly.

I agree that the situation with AF447 did involve pitot tube icing as a trigger to essentially unexplainable subsequent actions by the flight crew.

Perhaps the tragic Polish Air Force Tu-154 crash in 2010 in Smolensk, Russia is a better parallel:

http://en.wikipedia.org/wiki/2010_Polish_Air_Force_Tu-154_crash

It seems to me to be similar example of a 'do or die' approach (a go-around was not or became not an option), albeit the Tu-154 approach conditions were much more demanding (thick fog, around 500 meter visibility) than that experienced by the Asiana Airlines Flight 214 flight crew.

Edited July 9, 2013 by MaxYakov

[TallGuyJohninBKK]

I believe the NTSB has said they'll be checking the fuel system in the SFO crashed plane, to check for the same problem that brought down the plane at London years ago.

But, this plane had P&W engines, whereas the London 777 had Rolls Royce engines...and the problem isn't expected to repeat in the latest episode.

[EyesWideOpen]

The PAPI was working, which is just as accurate as a ILS glide slope, but neither would have been needed for this approach anyway.

Like AF 447 this accident appears to be another shocking example of airline pilots forgetting (or never having really learned) the basics when something goes wrong with the automation.

Massive difference between this accident and the Air France one. So big a difference I find it hard to draw a comparison with them as you have.

Flight was on final approach in clear weather operating under VFR. One of the fundamental basics of final approach under VMC is the positioning of the approach point through the cockpit windows, and adjusting your speed and attitude accordingly.

These accidents are so vastly different I still find it impossible to compare them.

A330-200 series vs Boeing 777.

One in the middle of the night, in pitch black conditions over the middle of the Atlantic Ocean at high altitude - the other in clear daylight in the middle of the day at ground level.

One involved the crew having obvious spatial disorientation and giving incorrect control inputs to match the actual state of flight - the other was a controlled flight into terrain.

The earlier post I responded to says they are comparing it to Air France 447 because it is another case of the pilots forgetting the basics when something goes wrong with the automation. I don't accept that premise for the 777 accident, which happened in clear daylight on final approach under full control.

Just my two cents worth.

You are missing the point. The position the aircraft is in is COMPLETELY irrelevant. For different reasons the aircraft were in a position of danger. In the case of the this plane, an improper approach. In the case of the Air France flight, it was a failure of the auto pilot link to the air speed indicator.

The Air France pilots did not have the skill to fly the plane at night without the autopilot, and the Asiana pilots did not have the skills to do a go around when they realized they were too low and slow. Sugar coat it all you want to try to reassure the flying public, these are pilot errors at the most basic level. The incompetence of the Air France pilots was shocking to say the least, and when the full report comes out on this crash I suspect it will be equally shocking.

If you want to see what real pilots do when the chips are down, read about BA Flight 9, the flight where they lost four engines after flying through a volcanic ash cloud.. I am pretty sure if the same thing happened to a Asian plane, everybody would have died.

http://en.wikipedia.org/wiki/British_Airways_Flight_9

Edited July 9, 2013 by EyesWideOpen

[NeverSure]

Either one [pilot] could have firewalled the throttles 30 seconds out and it would all have been a non-issue. That's assuming there was no mechanical failure.

According to Hersman's account I posted above, the plane was already below designed landing speed and on a slowing trajectory 34 seconds and 500 feet of altitude before impact.

So I guess the couple of hundreds of million dollar question is: Why didn't they -- prior to the 8 seconds before impact and 125 feet of altitude when throttles began moving forward?

I don't think we disagree.

30 seconds and 500 feet of altitude is still plenty of time and room to firewall the throttles and initiate a go-around. The landing should have been aborted, again assuming no mechanical problems.

The getting into position and stabilizing for landing starts way out and all is constantly monitored throughout descent. If things such as altitude or airspeed or descent rate (all of which can be monitored manually in that weather with no help from the airport's systems) are out of whack at 30 seconds and 500 feet, a full throttle go-around is THE answer.