Asiana 777 crashes in SFO

An Asiana Airlines Boeing 777 crash-landed at San Francisco Airport Saturday, killing at least two. This is the first fatal crash involving a 777.

Investigators will certainly look at whether fuel line icing may be a factor, which was traced to be the cause of the only other 777 accident, British Airways at London several years ago, also a crash-landing situation. Early news reports seem to reflect a similarity in the flight profile between the two flights. As readers know, we’re traveling and we don’t have access to our files to determine if Asiana uses Rolls-Royce engines, which are those used on BA and which were susceptible to icing.

GE engines on the 777-300ER have more recently come under some scrutiny for issues, and we’d expect investigators to consider whether there is any connection if Asiana uses GE on its 777-200s. This would be a natural course of considering all possible factors.

Other factors that will be looked at: human error, mechanical problems and controlled flight into terrain (CFIT).

Update, 5pm GMT: With the knowledge now that the engines are PW, fuel icing as a cause seems pretty unlikely, but CVR and FDR readouts will indicate engine performance parameters. Although weather doesn’t appear to be a factor, it will be evaluated for the prospect of any clear air windshear or other conditions that could be a contributing cause.

Statements by the airline officials at this point that there wasn’t any pilot error or mechanical issues are entirely premature, given when the statements were made the data recorders hadn’t been recovered much less read.

Closely looking at a photo seems to indicate the aft pressure bulkhead in place, meaning the tail severed aft it it.

66 Comments on “Asiana 777 crashes in SFO

    • already shows it as “written of”:
      Serial number 29171 LN:553
      Type 777-28EER
      First flight date 25/02/2006
      Engines 2 x PW PW4090

      Wonder where the fire(s) started? up from the hold?
      (obviously once you have a fire in the fuselage the crown
      will collect all the combustible gasses)

  1. Putting aside mechanical and similar technical issues, a likely cause is pilot error. Much of the high tech landing aids are out on that runway for upgrading, as well as recent changes to the thresholds. So good old fashioned visual piloting had to be done. It looks like the plane was a little high on the glideslope and they corrected for it, but ended up too slow. It was too late when they tried to recover and consequently hit the seawall at the end of the runway with the tail and landing gear. Contributing factors are claims about weaker CRM in Korean pilots (mostly former military) with some historical evidence, and lots of anecdotes (which are not data), as well as various modes on the cockpit aids and how they function without ILS.

    Of course that is currently indistinguishable from mechanical/technical issues, and it is only fair to assume the experienced crew were good pilots making the best of the situation they experienced. No matter what happened, virtually everyone walked away from it which is good.

    PPRUNE has good info on what was going on, details from current pilots who land 777s there, stereotyping and downright racism. Filter as you see fit.

  2. The pictures show the #1 engine (left), as well as the tail section missing from the wreckage. That engine could be under the left wing, but I suspect it is in the water, with the tail section and the main landing gear. The debris field (on land) starts at the seawall for Rwy 28L. The fact the airplane slid across the ground for only about 200 meters tells me the airplane was very slow (airspeed) at the point of collision with the seawall. This accident reminds me of DL-723, a DC-9 that hit the seawall for BOS Rwy 4R back in the early 1970s. DL-723 was flying the ILS in fog and rain at the time of that accident. In that accident there was initially one survivor, but he eventually died at a Boston Hospital sometime after the accident.

    Right now I will not speculate if the cause of OZ-214’s accident was a stall, weather, mechanical, or crew error. It is way to early to say anything. But flightaware is indicating the last radar speed reading was only 98 knots (ground speed) prior to impact. SFO winds at the time of the OZ accident was recorded as 210/07.

      • Detached starboard engine pushed into the fuselage, starting the fire — sounds plausible.

      • There seems to be no outside fire residue on the fuselage where that engine is
        is sitting.

        News here told about a (one of two dead?) victim having been run over by an ambulance.

        • Uwe, the report has been corrected to a fire truck may have run over one of the fatalities. That would make more sense as CFR part 139 requires all fire fighting trucks to be at the crash site within a few minutes. The ambulances would not have responded yet.

          It appears both fatalities were seated in or near the last row of seats, as both were found on the runway blast pad and, most probably were ejected from the airplane when the tail was ripped off. It was on the runway blast pad where the one victim may have been run over by the fire truck.

          My guess is, and this is only a guess, the fire truck driver was looking at the main wreckage to determine his best position to fight the fire and never saw that he may have run over someone. Since he ran over some of the airplane debris, a bump from a body may not have alerted him that he ran over one.

  3. Surprising to note the extent of the fire inside the cabin, given the fire retardant nature of aircraft furnishings you have to wonder what she had in the hold.

  4. From an Air Canada captain on

    Don, I’ve done the YVR-PEK-YVR route twice in the last two weeks and the upper levels temperatures have been quite warm. -37C to – 45C at FL370 up over Russia and through Alaska. Lowest temperatures on that route recently have only been around -55C at FL390, and it’s very rare to see a B777 at those altitudes on that route. So I think you’re correct that cold fuel temperature is an unlikely cause.

  5. In any case- the cabin attendants did a fantastic job getting almost everyone off ASAP. apparently all exits were opened – not too sure how many were used.

    Some claims were made that this was the first major accident at SFO. Splitting hairs- in 1970-71 a 747 on takeoff wiped out some approach lights, drove debris thru cabin and I seem to recall one person was killed.. One reason was a mixup of runways available at the time due to construction…


    Asiana CEO rules out mechanical failure in deadly crash

    excerpt ..

    …. It was the first fatal crash of a commercial airline in the U.S. since February 2009.

    The flight, which originated in Shanghai China before stopping in Seoul en route to San Francisco, was carrying 61 U.S. citizens, 77 South Koreans and 141 Chinese.

    Yoon Young-doo, the president and chief executive of the airline, speaking at company headquarters Sunday, said, “I bow my head and sincerely apologize for causing concern to the passengers, families and our people.”

    “For now, we acknowledge that there were no problems caused by the 777-200 plane or engines,” Yoon told reporters Sunday at the company headquarters.
    asiana crash

    A Boeing 777 airplane lies burned on the runway after it crash landed at San Francisco International Airport July 6, 2013.(Photo: Ezra Shaw, Getty Images)

    He declined to comment directly on whether the crash was due to pilot error or air-traffic controllers, but said the three captains on board had more than 10,000 flying hours of experience between them.

    Early Sunday, South Korea’s Ministry of Land, Infrastructure, and Transport identified the two pilots flying the jetliner at the time of the crash as Lee Jeong-min and Lee Gang-guk. The ministry said that four pilots were on board and rotated in two-person shifts during the ten-hour flight from Seoul.

  7. Uwe :
    During the evacuation the smoke seems to rise from behind the fuselage ( starboard side ):

    This picture, posted by my friend Uwe, seems to show discoloration on the upper fuselage, fore and aft. I am not sure at this point if this is the beginning sign of a fire inside and heating up the crown, or if it is dust that settled upon the wreckage after it stopped sliding.

    • My guess would go towards dust ( would fit in with the “dust plumes” seen on
      the long distance CCTV footage/captures shown ). You can see some swirls
      in the dust over the wing.
      Additionally passenger egress doesn’t appear panic driven and there is no smoke from
      the doors or the ripped fuselage crown.

      • ….Additionally passenger egress doesn’t appear panic driven and there is no smoke from
        the doors or the ripped fuselage crown.

        Ripped fuselage crown- ahead of aft pressure bulkhead ???? some pics do show a jagged tear in outer skin a foot or two ahead of aft pressure bulkhead – but until later with fire burn through rest of crown appears intact initially.

  8. Like the China Airlines 737 fire in 2007, there are pictures of people evacuating with their roll aboard bags. When they do evacuation tests for the 737 MAX and A320 NEO with higher capacities, they need to have every tenth person exit with a roll aboard bag to better simulate the behavior of the general public.

  9. I don’t comment on accident, but the pilots declared emergency didn’t they?

    Low fuel?

    • Apparently only after on ground as plane was stopping or stopped.

      • Don and TopBoom.

        You are right.
        There was no emergency before the aircraft hit the ground.

    • According to the tape of the local tower controller that I have heard, there was no emergency declared and nothing said about the fuel state of the aircraft.

      Now if they said those things before the time stamp I heard, or said something to the approach controller prior to hand off to the tower, I don’t know because I have not heard any of those tapes.

  10. Uwe, thanks for the very interesting photo: There appears to be a pair of compression cracks in the fuselage immediately above the leading edge of the port wing, and the crack then extends over the top of the fuselage, although there it appears to be somewhat crinkled but apparently not separated open, as could be expected if broken apart in tension. Unfortunately, the fuselage in that area may have melted away by fire subsequent to the photo.

    By aligning a strait edge along the bottom of the window openings above the port wing, it appears that the fuselage behind the crack is sloping slightly downward compared to the windows forward from the wing. The excellent structural integrity of the fuselage should be recognized, and perhaps some analysis of the crack failure can be useful for future designs.

    • Front fuselage seems to have been bent over to the portside relative to the center part of the fuselage. Stringer compression failure, one windows looks “folded” on that side.

      How much time between the “evacuation with photoshoot” and the arrival of emergency services to extract the heavily injured? Progress of fire into the cabin must have been
      even later.

  11. On further examination of the photo using a magnifier glass, it does appear that the skin at the top of the fuselage may be be cracked open, as could be expected if it had failed in tension. And there appears to be a tiny piece of skinmetal peeled back and sticking upward at the very top.

    Perhaps the tail section of the aircraft struck the bay headwall and was torn off in tension while the rest of the aircraft was angled upward with engines at high thrust in a futile attempt to gain altitude. This could have slammed the fuselage down hard on what may have remained of the landing gear causing tension at the top and compression at the bottom of the cylindrical fuselage.

    Credit is surely deserved for the soundness of the design and construction that made it possible for so many passengers to exit a crashed fuselage and walk away.

    • German late night news shew a complete video sequence of the (crash)landing
      from still in the air to the final movements. ( So it must be available elsewhere too )
      There seems to be some horizontal cartwheeling about half way along the path on land
      with one wing rather high, then in the final a big dust cloud.

      • A bit more – the video looks to ME that the tail may have hit the water ( with spray ) just before it hit the seawall… and the rest of cnn has this

        (CNN) — The cockpit voice recorder of Asiana Airlines Flight 214 appears to show the pilots attempting to abort the landing just 1.5 seconds before it crashed at San Francisco International Airport, the National Transportation Safety Board chairman said Sunday.

        The pilots appear to have increased speed 7 seconds before impact, and they then “called to initiate a go-around 1.5 seconds to impact,” Deborah Hersman said.

        The NTSB’s preliminary assessment of the plane’s cockpit and flight data recorders show the flight was coming in too slow and too low, but Hersman stopped short of pinning the blame on the pilots.

        “We have a long way to go in this investigation,” she said.

        Survivors of the crash were being treated Sunday for injuries ranging from paralysis to “severe road rash.”

        But they’re alive.

        In all, 182 people were hospitalized and 123 others walked away from Saturday’s crash landing of a Boeing 777. The number who emerged unscathed prompted the city’s fire chief to describe it as “nothing short of a miracle.”

        Amateur video surfaced on CNN Sunday showing Asiana Airlines Flight 214 approaching the runway and striking what appears to be a seawall before rotating counterclockwise and coming to a stop. Fred Hayes said he shot the video about a mile from the crash scene.

        The death toll remained unchanged Sunday. Two 16-year-old girls died in the crash.

        “We were expecting a lot of burns,” said Dr. Margaret Knudson, San Francisco General Hospital’s chief of surgery. “But we didn’t see them.”

        At San Francisco General, 19 survivors remained hospitalized, six of them in critical condition.
        Plane crash-lands in San Francisco Plane crash-lands in San Francisco
        Details of pilot flying at time of crash
        Passengers recount moment of crash
        Passengers describe harrowing crash
        ‘The wheels … were too low, too soon’

        Many of the injured said they were sitting toward the rear of the aircraft, said Knudson. Several suffered abdominal injuries and spine fractures, some of which include paralysis and head trauma, Knudson said. Many patients also were treated for “severe road rash,” she said, which suggests “that they were dragged.”

        The conditions of victims at other hospitals was unclear Sunday.

        In Washington, investigators were examining both flight data recorders, which could reveal clues to what caused the crash landing.

        Survivors and witnesses reported the 7-year-old airliner appeared to be flying too low as it approached the end of a runway near the bay.

        “Stabilized approaches have long been a safety concern for the aviation community,” Hersman told CNN on Sunday, saying they represent a significant threat. “We see a lot of runway crashes.”

        “We want to understand what was going on with this crew so we can learn from it,” Hersman said.

        Hersman said her team hopes to interview the pilots in the coming days.

  12. Sorry- I should have said passengers and crew, who did a very commendable job getting everyone who was ambulatory off promptly — even some with carry-off luggage which the rules require should be left behind so the evacuation can go more quickly.

  13. Comparing the flight tracking data on the flight that crashed ( with the equivalent data from the previous day’s flight (, the rate of descent in the final minutes seems larger on yesterday’s flight. There are differences in the altitude profile that make it difficult for me to compare these. Someone with relevant experience may see something significant in these tracking logs.

    Soon there will be more precise data available from the flight data recorder,

  14. Loooks like a too late realization of need to go around

    Associated Press

    PREV 1 of 5 NEXT


    This aerial photo shows the wreckage of the Asiana Flight 214 airplane after it crashed at the San Francisco International Airport Saturday morning.

    Boeing 777 jet known as one of the safest
    After S.F. crash, fliers at Sea-Tac face rebooking plans
    Gallery: Asiana 777 crash-lands at San Francisco airport
    A look back at previous airport approach crashes
    Two Dead, 181 Hurt in Crash

    Raw aerial footage of wreckage

    Asiana Airlines

    Asiana Airlines is South Korea’s second-largest airline. It has recently tried to expand its presence in the United States and joined the Star Alliance, which is anchored in the U.S. by United Airlines. A closer look:

    Founded: 1988

    Fleet: 79 planes, including 12 777-200ERs

    Headquarters: Seoul

    Daily flights: 268

    Destinations: 23 countries, 71 cities, including Seattle and New York, plus 12 cities in South Korea

    Employees: 9,073

    Asiana Airlines, Star Alliance

    Landed short. Question is why. No evidence of fire before landing? No mayday calls. … (July 6, 2013, by alfa12) MORE
    Tail-strike on landing, according to various sources. (July 6, 2013, by LurgidBee) MORE
    The Aircraft made a tailstrike against the seawall breaking off the empennage and the… (July 6, 2013, by kn0man) MORE
    Read all 81 comments Post a comment
    A federal safety official said the cockpit voice recorder from Asiana Airlines Flight 214 showed the jetliner tried to abort its landing and come around for another try 1.5 seconds before it crashed at San Francisco airport.

    National Transportation Safety Board chief Deborah Hersman said at a news conference Sunday the recorder also showed there was a call to increase airspeed roughly two seconds before impact.

    Before that, she said, there was no indication in the recordings that the aircraft was having any problems before it crashed Saturday, killing two passengers and injuring scores of others.

    Investigators took the flight data recorder to Washington, D.C., overnight to begin examining its contents for clues to the last moments of the flight, officials said. They also plan to interview the pilots, the crew and passengers.

    “I think we’re very thankful that the numbers were not worse when it came to fatalities and injuries,” Hersman told NBC’s “Meet the Press.” “It could have been much worse.”


  15. A brave initial concept combined with an creative design, holding an almost unblemished operational history that given the number of operators some of which have an accepted dubious maintenance & safety history, it’s nuts & bolts design ensures it maintains a safety record that is applauded within the industry & remains largely unchallenged.

    Preferring more established & historically reliable engine numbers (x4) on any long haul flight, in terms carrier selection on long haul it remains my last choice as does another twin.

    Having said that the two x four power plant debate does not appear to raise it’s head in this unfortunate incident.

    Other than stating the bleeding obvious it’s Interesting to see many frequent correspondents here have once again become self proclaimed experts seemingly fed by watching TV & being drawn in by the usual media frenzy caused by filling rubbish airtime.

    • This is a superb analysis, and the closing update with the NTSB update is crucial. In the acornym suggested by my wife, this appears to be PFU (Pilot F..k Up). CFIT.

      • I agree, it’s excellent, sadly hardy likely the media would have neither bothered to hunt down or understand the Flying Professor.

      • Yep, sure looks like it. It seems he was hosed by 4 nmi out, high sink rate, low engine power… bad place to be in.

      • In 1968 a Japan Air Lines DC-8 on approach to Runway 28L at SFO landed in the water:

        – At the NTSB hearing, Captain Asoh took the stand as first witness and supposedly said, in answer to why he had landed in the bay, “As you Americans say, I fucked up.”

        • One of the photos shows that the bay-end of that SFO runway has a very sharp edge — apparently sharp enough to shear of the 777’s entire tail assembly, leaving the aft bulkhead partially destroyed. This sort of mini-cliff runway-edge is unique at SFO, as most airport runways end fairly level with the surrounding surfaces. There may also be a unique visual depth perception problem because the runway is elevated only a few feet above the surface of the bay’s water.

          If a bunch of old automobile tires had been anchored down to form a buffer roughly level with the sharp vertical edge at the end of SFO’s concrete runway to soften that edge, would that tail assembly possibly have had a chance to remain intact, or at least have sustained less severe damage? Would such an inexpensive precaution possibly have reduced the severe injuries and two deaths suffered by passengers in the rear of that crashing aircraft?

          • SFO is not unique for having seawalls protect runways. BOS, HNL, and many more around the world also have seawall protected runways. Runways that end in sea water need to be elevated by a few feet to prevent flooding during extreme high tides. Large rip rap boulders need to be placed to control erosion caused by storms. They must be heavy and strong enough to take continuous pounding by the water. Don’t forget, SFO is in a heavy Earthquake Zone, and it would not be unusual for an Earthquake in the East Bay Area to send a small tsunami towards the SFO airport.

            SFO Rwy 28L had an operating PAPI (at least until the aircraft wreckage took it our), a displaced threshold, and remarked runway markings. But, for some unknown reason these visual nav-aids did not help OZ-214.

          • .OK, so why not have a bunch of old scrap tires lying flat and anchored to the rip-rap sloping up from those seawalls at all airports with seawalls at the end of their runways? Old worn-out scrap tires are very inexpensive — in fact, there are big repositories of them in many places that would be delighted to get rid of them by having some of them recycled elsewhere.

            TC has posted a similar idea, but scrap tires would be much less expensive and could be in place much more quickly than adding a 10-degree paved bank at existing runways that end in seawalls.

            Bob Paglee, P.E.,(Ret.)

  16. First, my condolences to the families of the two girls who lost their lives. Such a terrible tradgedy.

    I drove to SFO to see the wreckage-completely surreal.

    I was really surprised as how the fuselage stayed intact given how hard the fuselage got “slammed”-twice as far as I can see-once on impact and the second time when the plane “twisted”.

    The fact the fuselage stayed intact definitely contributed to the lack of fatalities or an increase in injuries.

    To me, ostensibly it seemed like pilot error. I guess we’ll find out soon enough.


  17. According to what I have read ILS was off.

    Radio communication:|tvideo|article

    Some good pictures I found here:

    On the 2nd picture you see the heck cone (with rudder?) on the upper left part of the picture and the two elevators on the right. The main landing gear seems to be disintegrated right there at quay wall. Some tires or axes could lie at the left elevator.
    On the 3rd picture the left engine can be seen on the far left side of the picture.

    Here is a video with several pictures:

    On the last picture you can see what it might be a bad idea to put passengers in the cargo compartment.

  18. Surprised that no one has suggested unreliable air speed indication.
    Quite sure that if it had been an A330 it would have reared its head.
    Hard to believe a flight deck crew collectively failed to see an ASI showing a seriously low approach speed.
    Lets not crucify these guys who must be feeling pretty miserable before we get the facts!

    • No, I don’t think you are correct in your assumption, at least as far as it being just two days after the accident (unless there was evidence to suspect an airspeed indicator/pitiot tube problem).

      The pilot was in training in the B-777, but had some 9700 flying hours in the A-320. This was his first landing attempt at SFO in the B-777.

      Here is a video tape someone shot showing the last 30 seconds, or so of the low, slow, unstabilized approach and into the crash sequence. Notice how high the nose is on the approach, the initial impact (partially obscured by the holding B-747), and near the end of the crash sequence the main portion of the wreck cartwheeling.

  19. Watching the video of the landing gave me the impression that hundreds would die. It is a miracle that only 2 did.

  20. BBC reports that: “Cabin manager Lee Yoon-hye told reporters that during the evacuation, two slides had inflated toward the inside of the plane, instead of outside, hurting two flight attendants.”
    CNN reports that one of the fatalities may have been caused by an emergency vehicle running over the passenger on the runway.
    Cases of road rash where other passengers were ejected.

    With all this it is very amazing only two died.

    • The B-777 seems to have a more advanced body than those airplanes designed just 10 years before the B-777 was. If we compare this accident to the 8U-771 accident at LLT in 2010, that was a modern airliner, too. That accident involved an almost brand new A-330, designed about 8-10 years before the B-777 was. In that accident, the airplane disintegrated upon ground impact and there was only 1 survivor, a Dutch boy, IIRC.

      The similarity between 8U-771 and OZ-214 is both airplanes became low and slow on the final approach, although the 8U flight did try to go around, the crew screwed it up and their aircraft crashed and disintegrated.

      Although both airplane had different dynamics to hitting the ground are different, and to my knowledge no tape exists showing the 8U crash, we do have various videos of the OZ crash. The tape shows a very violent twist and cartwheel of the B-777 as the airplane finally settled.

      So, Normand Hamel’s link to the Seattle Times story does make some sense. What the story cannot say is every accident is unique and engineers and designers cannot build any airplane that will guarantee 100% survival in every accident.

      The robustness of Airbus and Boeing modern airliners first showed these strengths in the QF-32 accident, an A-380 that suffered an uncontained engine explosion that took out most of the remaining aircraft systems. Fortunately for that incident there were 5 very experienced pilots aboard who were able to share responsibilities to get her safely on the ground. The redesigned B-737NG also showed these strengths in both the AA-331 and the JT-904 accidents

      • Circumstances for this and the Afriqiyah Airways Flight 771 are completely different.

        The A330 had been set to TOGA power and lost the last 180′ in 2 seconds that is a sinkrate of about 5400′ /minute. “normal” landing speed and accelerating.
        Do the math and you will get 55kn downward and 150..170kn forward impact speed.
        ( decelerating from 50kn in 3m (passenger feet) demands 150g, in 6m ( full fuselage height) 75g )

        Compare to an apporach @ ~105kn, very low sinkrate and probably flying in ground effect. Forward energy dissipated over 1000++m of soft earth. For most seats forces seem to have not gone significantly beyond the required 15g resillience.

  21. The amateur film clip shows the starboard wing high in the air while the aircraft spun through what looks like 400 degrees.
    The whole lower half of the aft pressure bulkhead is gone and several passengers say the aft lavs and galleys were gone and they exited the aircraft through that hole.
    So far as I know they still haven’t located the No. 1 engine; it is probably in the water either short of the seawall or in the inlet to the left of the wreckage.
    Hard to believe that it would be hidden under the fuselage, but stranger things have happened.
    I am surprised that there is no reported damage to the approach lights on the pier extending out from the runway 28L centerline.

  22. It apears that SFO’s lack of an instrument landing system also played a part in this incident, with pilots of the type that fly into this airport claiming “it was a question of time”. As it is often the case in accidents, it is a series of related chain of events that contribute to a catastrophe more than a single isolated cause. Check out:

  23. Where runways go into the water, they could design them with more forgiving angles and material. A 45 degree rip rap bank could be replaced with a 10 degree paved bank. Especially if the nose is at water level headed for the bank.

    Looks like a DC-8 landed short once too.

  24. On the 3rd picture the left engine can be seen on the far left side of the picture. You have to enlarge the picture to see it. There is a black rectangle on the left side of the picture. The yellow engine is left to it and just a little below.

    Yes, the 21st frame shows it clearly.
    Odd that the amateur video does not show it departing the aircraft during the ground loop.
    Must have been either too small to notice or obscured but the dust cloud.

  25. Ooops that was yesterday – but todays should be up soon

  26. Children of the magenta- regarding Otto mation

    about 25 minutes – good explanation of what probably happened re workload and basic abililty to fly and otto – throttle

Leave a Reply

Your email address will not be published. Required fields are marked *