Patterico's Pontifications

7/6/2013

Airplane Crash Lands in San Francisco

Filed under: General — Patterico @ 1:18 pm



This guy says he was on the flight, and that everyone is OK, as far as he can tell.

UPDATE: Unfortunately, reports now say there are two dead.

An Asiana Airlines flight from Seoul, South Korea, with more than 300 people aboard crashed while landing at San Francisco International Airport on Saturday, littering the runway with debris, and forcing passengers to jump down the emergency inflatable slides to safety.

San Francisco Fire Department told CBS News that two people were killed, with 61 injured, some critically, and that the number of injured “will go up.”

That’s too bad. I was (obviously) hoping everyone was OK.

102 Responses to “Airplane Crash Lands in San Francisco”

  1. Looks like it paid off to be in First Class.

    I wonder if it was a flight in from Asia… that’d suck. 14-16 hours of boredom. Then terror. Then burnt luggage.

    SteveG (794291)

  2. Holy crap! Watching on FNC. The roof is ripped off the jet. Foam all over the place. Been media-free until I turned on the set.

    L.N. Smithee (301382)

  3. Everything seems to be fine. So far, no word of deaths. Looks much worse than it was, apparently.

    L.N. Smithee (301382)

  4. OK, it looks like at least two people have died. Dammit.

    L.N. Smithee (301382)

  5. Staff back in the tail section?

    SteveG (794291)

  6. How horrible! I heard decades ago that the tail section was supposed to be the safest part of an airliner. I never understood why – maybe remoteness from the engines back when explosion on impact was common?

    felipe (6100bc)

  7. whatever the facts may be, I congratulate the pilots for minimizing the loss of life in landing.

    felipe (6100bc)

  8. I have to suggest that we congratulate the designers and engineers who, over the decades, have gotten aeroplanes to where this can happen and most folk literally can walk away safely …

    It also emphasises the differences between free enterprise and state-controlled development and control of airlines …

    Alastor (2e7f9f)

  9. SteveG-

    Flight 214 left Seoul’s Incheon International Airport earlier Saturday and flew 10 hours and 23 minutes to California, according to FlightAware, a website that offers tracking services for private and commercial air traffic. (CNN)

    elissa (ff048d)

  10. It looks like maybe the plane was too low and hit the end of the runway near the ocean.

    Every time I fly in there I wonder that it doesn’t happen more often. And pray, too.

    Patricia (be0117)

  11. Patricia–I agree. While SF is not my scariest airport to land at, it is high on my list.

    elissa (ff048d)

  12. Boeing, or AirBus?

    askeptic (2bb434)

  13. The media’s abject refusal to tamp down speculations, despite overwhelming evidence, is farce.

    The weather was near-perfect. That data was easily available and the media could not stop talking about possible wind shear.

    Pics showed many pax leaving the plane, with zero burn marks on the fuselage, yet anchors refused to acknowledge this.

    The debris trail was obvious from the get-go – that plane hit land right at the edge – not on the runway, yet it took them an hour to state this and instead kept talking about other possibilities.

    Asiana reported that these flights are operated with fewer than 310 souls on board. Yet, all we heard was estimates up to 400 or so.

    Anchors kept reporting that the plane was “reported to have flipped upside down” despite the live video showing no such thing. What was the point of using that false eyewitness report?

    The ATC audio was available through twitter no later than 45 minutes after the crash. Why not have one of these pilots talk through the ATC audio – which clearly shows no distress until after the crash?

    Any chance George Zimmerman was on board and caused this?

    Ed from SFV (6382f3)

  14. Boeing 777

    elissa (ff048d)

  15. Terrible. Is it clearer what happened now?

    Sarahw (b0e533)

  16. Hugh Hewitt tweeting that a pilot (and friend) who is a disaster litigator says the odds of a computer malfunction are extremely low and that this is almost certainly pilot error.

    Bingo.

    Ed from SFV (6382f3)

  17. It’s not the first Boeing that has lost a tail due to a hard landing.
    Seem to recall a 747 in Japan that had bumped its tail a time or two, and finally had a bulkhead separation; unfortunately at altitude.

    askeptic (2bb434)

  18. Sorry for the remark about burnt luggage. The passengers leaving all look pretty good in the photo and I incorrectly assumed everyone was more or less OK.
    Apologies to their loved ones

    SteveG (794291)

  19. I’ve flown Korean Air in Asia and I really like the service they provide.

    But based upon what I’ve been told about Korean airlines in general I’d never fly with them across the Pacific. This crash fits the pattern.

    Korean airlines don’t put relief crews aboard their transoceanic flights. When they crash, as in the case here, they do so at the end of long flights. The crew is tired and they tend to make mistakes. Also because of the culture the co-pilot doesn’t correct the captain even if the co-pilot is aware the captain is making mistakes. The junior person simply does not correct the senior person, whereas in a US airline it’s the job of the co-pilot to keep the captain safe.

    The commercial pilots I know simply will not allow their family members to fly Korean airlines. This Asiana crash illustrates why.

    Steve57 (c74c87)

  20. According to Asiana Airlines, 141 of the passengers who were aboard Flight 214 are Chinese, 77 are South Korean and 61 are American. (CNN)

    elissa (ff048d)

  21. Quit right, Alaster, quite right! The engineers and designers (and with all the improvements in aviation)are to be congratulated – even if pilot error was directly responsible for the accident, I still give them credit for bringing so many through it. No doubt the pilots and crew will be pilloried by someone -I wonder who.

    felipe (6100bc)

  22. Oops, I mean Alastor! Sorry, brother!

    felipe (6100bc)

  23. Sequestration ! Or whatever.

    Elephant Stone (00fc2d)

  24. Make it foolproof and God will make better fools.

    My condolences and sympathy to the victims, families, loving ones.

    htom (412a17)

  25. Just don’t like that Boeing 777. Or of course the 737. I won’t fly either, ever.

    Kevin Stafford (1d1b9e)

  26. 24. …My condolences and sympathy to the victims, families, loving ones.

    Comment by htom (412a17) — 7/6/2013 @ 5:48 pm

    Mine as well. I hope my critique of how Korean airlines operate doesn’t come across as lack of sympathy for the people killed or hurt in this crash.

    Steve57 (c74c87)

  27. . I hope my critique of how Korean airlines operate doesn’t come across as lack of sympathy for the people killed or hurt in this crash.

    You know, I lived in Korea for a while and I’ve flown on their planes a few times and I never experienced the problems you are citing. What’s your source of experience with Korean flight operations? At the end of the flight, the crew seemed quite fine to me, and they always had relief crews (aka they took breaks and other stewards worked). That is obviously also the case for the pilots.

    I would be shocked if the pilots didn’t correct eachother when they were making serious mistakes, and I will be surprised if it turns out that was the cause of this crash.

    I was much happier with Korean Air than with Delta. Korean culture is awesome, IMO, not deadly as is your opinion.

    Dustin (303dca)

  28. This crash has some similarities to the 747 crash in Afghanistan in which the plane stalled. That plane had more altitude and may have had a cargo shift but the same thing may have happened. In this case, the pilot realized , at the last minute, he was too low and pulled up too sharply. The fact that the plane stopped pretty quickly on the runway suggests a stall. I agree crew fatigue may have been a factor. SFO has a kind of scary approach.

    Mike K (dc6ffe)

  29. Re #10 Patricia: Every time I fly in there I wonder that it doesn’t happen more often. And pray, too.

    A commercial aircraft usually flies a “stabilized approach” that puts it over the runway in the right place at the right time. On a clear day the pilots do this visually and usually track along with their instruments as well. On a day with poor visibility they use the Instrument Landing System to provide visual cues in both azimuth (left-right) and elevation (up-down).

    It’s premature to speculate on causes at this time. It could be caused by pilot error (hard tail strike), by mechanical failure, or by some complicated operational reason. The investigation will take place in a professional and non-adversarial environment.

    Meanwhile, media figures and plaintiff’s attorneys will be conducting the air-crash equivalent of flinging poo. For example, CBS is now leading with the shocking news that this was a “survivable accident.” (Since there are many survivors, whoever at CBS thought this headline worthy may be brain-dead).

    This is the only official release to date:

    http://www.ntsb.gov/news/2013/130706.html

    However, NTSB has held a press conference where Debra Hersman spoke. I recommend giving full credence to her actual statements, exercising caution with paraphrases, and blowing off talking-head speculation entirely.

    Kevin R.C. O'Brien (2165ef)

  30. Re: Korean airline safety.

    This WSJ story, date uncertain, describes some of the problems. Long a subject of discussion in the industry. Note that it refers to KAL, not Asiana. This is background and is likely to have NO relation to the mishap in SFO.

    http://online.wsj.com/article/SB933090613281869060-search.html

    Kevin R.C. O'Brien (2165ef)

  31. Press conference from yesterday, San Francisco authorities only. also shows overhead of crash scene.

    http://www.youtube.com/watch?v=KaVlGEjjsL8

    At 6:30 in the video the conference is over and the news readers start speculating and filling dead air.

    There are two interesting factual items, one in the video, and one in claims I’m seeing in the news.

    1. The debris field begins before the runway, and the plane should not have touched down even on that early part of the runway — it’s a “displaced threshold” and normally a jetliner shoots to touch down 1000 ft/300m past the displaced threshold, in the case of this runway, 1300′ from the start of the pavement.

    Even the three separated tail fins and their control surfaces came to a halt short of the displaced threshold. So, while we don’t know why, we can definitely say the a/c touched down short of its proper landing point. It should have touched down 1300′ beyond the beginning of the pavement, and appears to have touched down before the beginning of the pavement. The pavement here projects out into the bay — here’s some information and there’s a downloadable diagram of the runways.

    http://airnav.com/airport/KSFO

    2. According to news stories, Ms. Hersman has said that the glide slope (the component the Instrument Landing System that provides vertical guidance) was out of service and the pilots had to land visually. I have not seen that in her own words, but note that the glide slope for RWY 28L was indeed NOTAMed out at the time. Some sources are claiming the visual approach slope lights, Precision Approach Path Indicators were NOTAMed out but they were not until afterword (the crashing plane took them out).

    3. I’ve listened to the tower audio and two things struck me. One is that the English of the pilot not flying was pretty awful; this is fairly common with Asian lines (and some others) but is unlikely to be a factor in this mishap. Another is there’s absolutely no indication of trouble. Only afterwards is the pilot not flying excited; he went from relaxed and confident to stressed-out very quickly, and in his voice, it’s only obvious after impact. The tower controller tells him the equipment is already rolling.

    4. The FBI is on scene but does not expect to take authority for the investigation.

    5. Given those facts, the investigation will probably be able to produce a defensible statement of probable cause in due course.

    Kevin R.C. O'Brien (2165ef)

  32. That’s too bad. I was (obviously) hoping everyone was OK.

    One hopes, but “only” two dead from an airplane crash, that’s not a miracle but it’ll play one on a Lifetime movie of the week…

    IGotBupkis, "Faeces Evenio, Mr. Holder?" (a2f645)

  33. (Since there are many survivors, whoever at CBS thought this headline worthy may be brain-dead).

    I believe this sentence is overlong and overly complicated. Allow me to suggest a more terse and clearly equally accurate variation for you, for consideration of future use:

    The journalists, editors, and talking heads at CBS are, as usual, brain dead.

    See? Much more straightforward and without all the hemming and hawing about known, observable data…
    😀

    IGotBupkis, "Faeces Evenio, Mr. Holder?" (a2f645)

  34. A FWIW comment-

    Of course every plane crash is a tragedy, but the fact that most of us still fly in them indicates how “safe” it is, as life goes.
    Years ago there was an article in the New England Journal of Medicine (NEJM- perhaps the most elite medical journal) that discussed the safety of the airline industry and commented on it being a culture of honesty and cooperation. The worst thing one could ever do was not be truthful about problems or potential problems. The more standard operating procedure of trying to minimize and shift liability, etc. that exists in the rest of the world was purposefully avoided in the airline industry.
    At least that was my understanding of it, when it was written 20 some years ago or so.
    One implication was that American society as a whole prefers wealthy lawyers and victims to a safe medical/healthcare system, where the pressures concerning litigation are opposite.

    People working together for the common good instead of adversarial for individual gain; sometimes it is a good thing.

    MD in Philly (3d3f72)

  35. Looks like a possible runway undershoot.

    When a plane gets low and slow and the pilot pitches the nose up without additional power the plane sinks faster. Aerodynamically, this is referred to as the “region of reverse command”

    From the video of the actual crash, the plane was coming in slow and with a nose high attitude. A plane in that condition requires a great amount of engine thrust to maintain altitude and even more to climb. Despite that, when coming into land at too low an altitude, the instinct for the pilot is to pull back on the yoke, which exacerbates the problem.

    Calfed (5b899d)

  36. Latest I heard is that records show there was an alarm for a possible stall with too slow of speed, and they requested permission to skip the landing and make another pass.

    No idea what that all means other than they knew there was a problem moments before the crash and were trying to do something to avert it
    but maybe as Calfed said, in their desire/attempt to miss the landing and try again they drove the tail down in the attempt to make the plane go up

    if that doesn’t make sense, ignore it

    MD in Philly (3d3f72)

  37. That does seem odd, MD, the deceleration would seem to be on approach.

    narciso (3fec35)

  38. Here’s some substantiation, MD. Somewhere, I think it was CNN, a survivor who was sitting near an engine said there was no emergency warning announcement in the cabin and he had not been paying any attention to the landing at all. Everything seemed normal– but then just before/as the plane crashed the engines revved loudly and mightily as tho the pilot were trying to rise rather than land.

    elissa (95fb63)

  39. That definitely sounds like the pilot realized that he was going to land short, tried to go around at the last minute and was unable to arrest the aircraft’s descent in time.

    Jet engines have to “spool up” before they provide thrust and there can be considerable lag between advancing the thrust levers and actually getting thrust out of the engines.

    As I said, however, in the configuration that the aircraft appeared to be in just before it hit, considerable thrust is required to maintain level flight, let alone climb.

    Calfed (5b899d)

  40. I don’t know if this information is reliable but according to the Aviation Safety Network:

    The weather at San Francisco was fine with 6-7 knot winds and a visibility of 10+ miles. The flight was cleared for an approach to runway 28L, the ILS glidepath of which had been declared unserviceable in the current Notam.

    The airplane was configured for landing with 30 degrees of flaps and gear down. Target approach speed was 137 knots. According to preliminary information from the cockpit voice recorder, the crew did not state and anomalies or concerns during the approach. The throttles were at idle and the airspeed slowed below target approach speed during the approach. Seven seconds prior to impact, one of the crew members made a call to increase speed. The stick shaker sounded 4 seconds prior to impact. One of the crew members made a call for go a around at 1.5 seconds before impact. The throttles were advanced and the engines appeared to respond normally. The rear fuselage then struck a sea wall, just short of runway 28L.
    ***
    The ILS glidepath for runway 28L and 28R at SFO had been declared unserviceable from June 1 until August 22.

    I believe the ILS glidepath is a reference to the Instrument Landing System that enables appropriately-equipped planes use autopilot to aid in landing. I wonder how many times these pilots had landed without using ILS.

    DRJ (a83b8b)

  41. A close friend who is a veteran commercial pilot for a major U.S. airline told me today in an email that the 777 is not considered to be a problematic airplane within the industry and that these are the first fatalities ever recorded from an accident with that aircraft model. He has not flown a 777 himself. He did not care to speculate on what may have happened in SF.

    elissa (95fb63)

  42. Interesting DRJ– I had seen some other blurbs about the landing system being off or disabled on some runways but not in so much detail as you have provided. I don’t know who or what entity manages this airport. Does anyone here know? Can lawsuits be far behind? I’m sure most commercial pilots would say that a fellow pilot of that category and responsibility should be able to handle a landing with little or no guidance or technology in a failure or emergency situation. But your question is a good one.

    elissa (95fb63)

  43. Another plane crash in Alaska, 10 presumed dead.

    JD (b63a52)

  44. Just for information, the “stick shaker” is a physical aid that lets the pilot know that the aircraft is approaching an aerodynamic stall. The control yoke actually starts shaking, as if to say “Hey, pilot, wake up”. This is another indication that the aircraft’s wings were approaching a critically high angle of attack.

    If the Aviation Safety Network information is correct, the aircraft was aerodynamically “dirty” (lots of drag), nose high, engines at idle (producing little or no thrust) and at the last minute the pilot initiated a “go around”, which means that he was attempting to climb back up to pattern altitude to come around and take another crack at the landing.

    Calfed (5b899d)

  45. The glideslope provides useful descent information to the pilot when it is operating, even if he chooses not to couple the autopilot to it.

    However, the weather looked clear in the video and making a visual approach should not have been a problem for any competent pilot.

    Calfed (5b899d)

  46. CNN has video of the crash

    DRJ (a83b8b)

  47. The CNN link says the pilot sitting in the captain’s seat had 43 hours of experience in this type aircraft. I’m sure he was an experienced pilot but it doesn’t sound like he was that experienced in this type aircraft.

    DRJ (a83b8b)

  48. Also from the CNN link:

    The NTSB has ruled out weather as a problem and said that conditions were right for a “visual landing.”

    But investigators are looking into whether construction at the airport may have played a role.

    Construction to extend a runway safety area temporarily shut off the so-called glide slope system, which is one of several options pilots have to help them land planes safely, Hersman said.

    DRJ (a83b8b)

  49. The glideslope is part of the Instrument Landing System (ILS) and is a beacon that sends a signal that is linear and aimed 3 degrees above the center line of the runway. It is necessary for making a precision approach in bad weather, but not necessary during visual meteorological conditions.

    It can be used, however, during a visual approach if it is in service. Once the aircraft intercepts the glideslope, it can be used to maintain a 3 degree descent, which will keep the pilot from getting too low or too high as he approaches the runway.

    Calfed (5b899d)

  50. In general, those who study disasters like the Space Shuttle say that there are usually a series if miscues that “line up perfectly” “like overlapping holes in Swiss Cheese”,
    so, as a total guess as an example that may mean nothing:
    -the glide slope system was off, which should not have been a problem, but perhaps was a surprise to a pilot not accustomed to the airport (?)
    -a pilot who was experienced, but not so much with this particular aircraft, hence the automated system would have been a help
    -other yet unknown factors which narrowed the safety margin for a safe landing

    with the possible result being no one specific thing causing the crash, but the overlap of several things, any one or more of which by themselves would not have caused the crash

    but as I said, that is all a general understanding of such things, there may be a very specific cause that will be identified

    MD in Philly (3d3f72)

  51. That is a good point, MD

    In aviation parlance, it is sometimes referred to as “the accident chain”. If one link had broken, the mishap would not have occurred.

    Calfed (5b899d)

  52. WAPO is claiming that according to Asiana this pilot had 10,000 hours of flight, but only 43 hours flying a 777 (as DRJ posted above) and that this was his first landing at SFO. The airline spokeswoman stated there was nothing mechanical wrong with the plane. Looks like they’re already going strong with pilot error and they’re ready to hang the pilot out to dry. But doesn’t the airline bear some responsibility for assigning a pilot with such a thin resume on this aircraft and airport?

    elissa (95fb63)

  53. So far it looks like pilot error. #40, Calfed @7:48 pm put the pieces together quite nicely.

    ropelight (20622b)

  54. We have a lot more clarity on the accident now, but NTSB will interview the crew. It’s doubtful the data recorders will tell us much, although it’s possible they may reveal that the crew did something blockheaded, like couple the autopilot to the dead glideslope. (In fact, KAL stuffed a plane in Guam some years ago “flying” a glideslope that was out of service, so as strange as it sounds, it can happen).

    It does look like pure pilot error at this time. Airplanes are a lot more reliable than humans. But we still won’t be sure until the NTSB releases its factual report in a few months. For the next few weeks the investigators will be learning things we won’t necessarily hear about.

    A lot of the things the media are talking about are red herrings. Pilot time in type is not a factor. That it was his first approach to SFO is not a factor. (A visual approach is a visual approach; the visual cues are the same in Anchorage and Zanzibar). And the airplane set-up is the same anywhere, visual or instrument.

    One trick some pilots use when a precision approach is out (i.e. no glideslope) is to set one of the two redundant navigation units to use a GPS-based VNAV approach. This is not standard procedure in any airline AFAIK, and airlines really discourage deviation from standard procedures. Doing the same thing the same way every time is one reason that crashes are extremely rare and we sometimes go years without losing a life in an airliner crash in the USA.

    That’s it for tonight… I have to save some energy for “hating cops.”

    Kevin R.C. O'Brien (2165ef)

  55. The video CNN has that DRJ linked to above is going to be priceless to everyone analyzing this.

    htom (412a17)

  56. The sequence of events leading to the crash begins with a lack of airspeed, the plane was coming in way too slow which resulted in the inability to maintain a proper descent angle.

    By the time the flight crew starts to take corrective action it’s already too late. Lack of airspeed translates to lack of options for recovery. As a last resort the pilot yanks back on the controls putting the aircraft into a severe nose-up configuration which initiates a stall exacerbating an already unrecoverable situation.

    Too slow, too low, and in a stall, the plane fails to clear the seawall, the empennage strikes the seawall at the aft bulkhead and breaks off flipping the 777 down hard on its nose and the crash continues as the plane skids down the runway.

    ropelight (3a876d)

  57. Actually, yanking back on the controls probably saved lives. By forcing the nose up in the last few seconds it prevented the plane’s mid section from striking the seawall which would have ripped the belly open and almost certainly killed everyone onboard.

    ropelight (3a876d)

  58. That it was his first approach to SFO is not a factor. (A visual approach is a visual approach; the visual cues are the same in Anchorage and Zanzibar).

    This is not exactly correct. The approach to runway 28L at San Francisco is an over water approach, which involves the lack of some visual cues which are present in an approach that is over land. Over water approaches can create the illusion that the plane is higher than it actually is, resulting in the pilot flying a lower than safe approach.

    It will be interesting to find out if the PAPI (Precision Approach Path Indicator) was out of service for runway 28L.

    The PAPI is a set of sights which helps the pilot determine whether he is high or low on the glide path in visual conditions. If it was out of service, that could be another link in the “accident chain”.

    Calfed (5b899d)

  59. I should have written…

    The PAPI is a set of lights which helps the pilot determine whether he is high or low on the glide path in visual conditions. If it was out of service, that could be another link in the “accident chain”.

    Calfed (5b899d)

  60. }}} it doesn’t sound like he was that experienced in this type aircraft.

    Yeah, well, I strongly suspect he isn’t going to get too much more… :-9

    IGotBupkis, "Faeces Evenio, Mr. Holder?" (a2f645)

  61. 55 Comment by Kevin R.C. O’Brien (2165ef) — 7/7/2013 @ 11:06 pm

    We have a lot more clarity on the accident now,

    The NTSB concluded its press conference just before the 6 pm news on Sunday.

    A lot of the things the media are talking about are red herrings. Pilot time in type is not a factor. That it was his first approach to SFO is not a factor. (A visual approach is a visual approach; the visual cues are the same in Anchorage and Zanzibar). And the airplane set-up is the same anywhere, visual or instrument.

    This was not only his first approach to SFO, it was his first time landing of a Boeing 777, ever.

    All this has to be have been a factor. Why now?

    Doing the same thing the same way every time is one reason that crashes are extremely rare and we sometimes go years without losing a life in an airliner crash in the USA.

    This was the first time, so how could he be doing the same thing the same way?

    San Francisco airport is considered one of the more difficult landing approaches, on a par with LaGuardia Airport in New York City, but not as difficult as Reagan National Airport in Washington, D.C.

    One thing is true – you can make up for imperfections in the airplane by better training, and knowledge of the machine, and any kind of a plane requires skill.

    Sammy Finkelman (d22d64)

  62. When the air speed indicator kicked in, he tried to pull up, but he hit a wall.

    One of the passengers killed was probably killed by a fire truck after she got out of the plane.

    Other asecvts of the rescue worked better.

    Sammy Finkelman (d22d64)

  63. Sammy, I’m amazed at the rescue, particularly the crew of the aircraft, who managed to get everyone off that burning wreck so quickly.

    You may be right that responders caused one death, which is sad if so, but one of those things that happens in a chaotic situations where time is running out.

    It does sound like the pilot needed more experience with this aircraft (I imagine that is why he was flying it and getting that experience). I still object to the claim that it was because he was Korean that he crashed (some nonsense about Koreans not correcting eachother when their elders make mistakes).

    Can a pilot request ILS be turned on, or is the process more complicated than flipping a switch? Why would it ever be off, even on a clear day, if this is a difficult landing?

    Dustin (303dca)

  64. I can’t recall so many survivors from a crash of this type before. Maybe Boeing really is building better airframes.

    SPQR (768505)

  65. SPQR, it’s miraculous. There are obviously a lot of ‘what went wrongs’ worth looking into, but there are also some ‘what went rights’ that deserve a mention.

    I do think one aspect is that this was the end of the flight (not much fuel left).

    Dustin (303dca)

  66. Can a pilot request ILS be turned on, or is the process more complicated than flipping a switch? Why would it ever be off, even on a clear day, if this is a difficult landing?

    The ILS is usually in operation even when the conditions are VFR. I believe the ILS has been out of service for Rwy 28L for sometime due to construction at the approach end of the runway.

    Most runways at San Francisco have either a PAPI or VASI (visual approach slope indicator), which provides visual descent information, kind of like the “meatball” on an aircraft carrier. I do not know if the PAPI for Rwy 28L was in service at the time of the mishap.

    Calfed (5b899d)

  67. I can’t recall so many survivors from a crash of this type before. Maybe Boeing really is building better airframes.

    One of the passengers was interviewed and indicated that the hole made in the fuselage when the tail came off allowed many of the passengers near the back of the plane to get out quickly.

    Calfed (5b899d)

  68. The seatbelts and seats were made to withstand 16 Gs. Some people had to be cut out of their seat belts. The crew had some equipment and also got knives from the police.

    Another thing that is said to have gone right is triage.

    Sammy Finkelman (a4dbab)

  69. The fire did not start immediately or at least did not get big.

    Sammy Finkelman (a4dbab)

  70. The PAPI was in service per NTSB. It was notamed out after the accident (which is not really meaningful as 28L is still closed, but if you check the timestamp on the NOTAM there it is). Ironically, it had been out for a few weeks for construction, because new pads were poured and it was relocated to the new displaced threshold, 300′ from the original. This was done on 28L and 28R. The 28R PAPI remains up.

    Call to increase speed (which was “significantly” [NTSB’s word] below bug all the way down) came 7 sec pre-impact. Stick shaker 4 sec pre-impact. Call to go-around came 1-2 secs pre-impact, again, per NTSB. Engines then spooled up nominally. CVR and FDR data all good.

    The first time thing should not matter. The pilot flying was reportedly a captain on 74s who had landed the 777 simulator over and over and over again. The pilot not flying was a training captain, supposed to be monitoring him. Landing 1300+ feet short of the touchdown point is not a consequence of a botched type transition, it’s a failure of basic private pilot level airmanship. Why the training captain didn’t correct, is unclear.

    I dunno the 772s automation, but wonder if they were following a phantom glideslope and ignoring the OOS flag?

    People survived because:
    1. like most widebodies the 777 is built like a brick lavatory.

    2. The hit was hard but not unsurvivable. If they had been 3, 5, 10′ lower at the threshold it might have been a real nightmare.

    3. The fire was slow to start. The wing and centerline fuel tanks probably did not rupture (or the centerline wing tank may have been empty and nitrogen-inerted). It looks like the fire came from the #1 engine which was alongside the fuselage.

    4. People wasted no time getting their heinies off the plane. Other pax and the flight attendants helped the injured.

    I’ve commented many times about inaccuracy of early reports in fast-breaking news. For example, even the SF politicians and chiefs didn’t know where sixty-something pax were at one point. Turns out, they were OK, but had been brought to the wrong assembly point and so didn’t get counted when they counted noses.

    Running over a body was a self-report from a driver. It is likely that person, who was in the debris field and not obvious from the driver’s seat of the fire truck, was already dead.

    Kevin R.C. O'Brien (2165ef)

  71. Correction, engine alongside fuselage was #2.

    NTSB is tweeting information, pictures and links to videos.

    https://twitter.com/NTSB

    Kevin R.C. O'Brien (2165ef)

  72. The ILS is usually in operation even when the conditions are VFR. I believe the ILS has been out of service for Rwy 28L for sometime due to construction at the approach end of the runway.

    Most runways at San Francisco have either a PAPI or VASI (visual approach slope indicator), which provides visual descent information, kind of like the “meatball” on an aircraft carrier. I do not know if the PAPI for Rwy 28L was in service at the time of the mishap.

    Comment by Calfed (5b899d) — 7/8/2013

    I see. Thanks for the explanation. At least I know it wasn’t an easy switch to flip, but construction.

    Dustin (0960a8)

  73. THE CBS EVening News (and others) reported today that the NTSB has interviewed the 4 pilots (3 of whom were in the flight deck) and they say the auto throttle hasd been engaged and was supposed to have set the speed at 137 knots.

    That plane didn’t have a co-pilot on the right seat but an certified instructor pilot. Howeer, it wss his first time acting as an instructor pilot with that pilot. He’d been certified as a flight instructior for a month.

    He is supposed to take over in emergencies, but even at the end here was giving instructions to the pilot being trained. Anyway, none of them had a clue that anything was wrong, that the plane was going too slow for the approach, for the runway, until 7 seconds before it crashed.

    2 flight attendants were ejected from the plane when it crashed, but were unharmed.

    Sammy Finkelman (a4dbab)

  74. So the auto throttle was broken? Seems like speed and altitude are things you don’t just rely on cruise control to monitor as you land, so this isn’t a good excuse.

    Dustin (dfe6b6)

  75. “2 flight attendants were ejected from the plane when it crashed, but were unharmed”

    wow.

    felipe (6100bc)

  76. I’m guessing that NTSB will figure out the question of the auto-throttle.

    felipe (6100bc)

  77. Apparently there were four pilots on board, two to rest while two piloted. That should put to rest the ‘It’s because they were koreans who don’t have relief pilots’ BS.

    Dustin (dfe6b6)

  78. [dumb question] There have been a couple of mentions of using a GPS-based artificial glide-slope substitute. Could they have entered the end of that structure out in the bay, rather than the desired touchdown point? [/dumb question]

    htom (412a17)

  79. 75. Comment by Dustin (dfe6b6) — 7/9/2013 @ 5:00 pm

    So the auto throttle was broken?

    Maybe it wasn’t.

    It hit me.

    It could be they entered the code for the wrong airport. Or maybe they gave the autopilot the wrong GPS.

    Seems like speed and altitude are things you don’t just rely on cruise control to monitor as you land,

    http://www.airliners.net/aviation-forums/tech_ops/read.main/111845/

    so this isn’t a good excuse.

    Maybe instrument landing is all they know, and they can’t do a visual landing.

    I think a lot of that happens now. And now we learn:

    http://www.myfoxtwincities.com/story/22798901/ntsb-autothrottle-was-not-engaged-as-asiana-pilots-thought

    “An overreliance on automated cockpit systems has figured in dozens of air crashes and incidents in recent years.”

    Sammy Finkelman (c4d9d6)

  80. 76. Actually the two fight attendants were injured, but not very seriously.

    Sammy Finkelman (c4d9d6)

  81. http://www.npr.org/templates/story/story.php?storyId=200513147

    National Transportation Safety Board chairman Deborah Hersman said the autothrottle was set for 157 mph and the pilots assumed it was controlling the plane’s airspeed. However, the autothrottle was only “armed,” she said.

    Hersman said the pilot at the controls, identified by Korean authorities as Lee Gang-guk, was only about halfway through his training on the Boeing 777 and was landing that type of aircraft at the San Francisco airport for the first time ever. And the co-pilot, identified as Lee Jeong-Min, was on his first trip as a flight instructor.

    Autothrottles typically have three settings — off, on and armed. An autothrottle that is armed but not on will remain at its previous speed, which was probably near idle, said Doug Moss, a pilot for a major U.S. airline and an aviation safety consultant in Torrance, Calif. Pilots will frequently shift to idle off and on when preparing to land in order to descend faster.

    The pilot flying the plane had turned off his flight director, while the training captain had his flight director on, Hersman said. The flight director computes and displays the proper pitch and bank angles required in order for the aircraft to follow a selected path.

    In most airliners, an autothrottle will not turn on if one flight director is off and one on because it has to work in harmony with the flight directors — both need to be either on or off, Moss said.

    Only moments before the crash did the training captain realize the autothrottle wasn’t controlling the plane’s speed, Hersman said.

    “This is one of the two hallmarks of complexity and challenge in the industry right now,” Moss said. “It’s automation confusion because from what Deborah Hersman said, it appears very likely the pilots were confused as to what autothrottle and pitch mode the airplane was in. It’s very likely they believed the autothrottles were on when in fact they were only armed.”

    Sammy Finkelman (c4d9d6)

  82. They didn’t know how the equipment worked.

    Sammy Finkelman (c4d9d6)

  83. There have been a couple of mentions of using a GPS-based artificial glide-slope substitute. Could they have entered the end of that structure out in the bay, rather than the desired touchdown point?

    I don’t think that is likely. The approaches are contained in a data-base within the GPS unit itself and are not generally entered manually by the user.

    In fact, there is a GPS approach to runway 28L at San Francisco which has very low minimums that the pilots could have flown had they chosen to.

    My guess is that the pilots requested and were cleared for a visual approach and flew the approach visually. Why they allowed themselves to get so low and slow is the real mystery here.

    Calfed (5b899d)

  84. I guess I just assume that there are analog speedometers and altimeters on board, and that pilots glance at them regularly when they are landing, just because, hey, a lot is at stake with these computers running the show. They may not know the precise perfect speed, but they surely have a good idea of when the speed is way too low.

    Anyway, your quotes about the way automation complexifies are very interesting and may turn out to be highly relevant.

    Dustin (dfe6b6)

  85. Apparently the plane was coming in too slow, and this was noticed, and the autothrottle was turned on in an attempt to set the speed.

    But the instructor pilot didn’t know it wouldn’t work unless both flight directors were turned on.

    Sammy Finkelman (d22d64)

  86. Last word.

    peedoffamerican (ee1de0)

  87. Nope.

    SPQR (768505)

  88. I see what you did there.

    peedoffamerican (ee1de0)

  89. Did you see what I did?

    peedoffamerican (ee1de0)

  90. The pilot claims he was blinded by a bright light at about 500 feet, about the time they realized the plane was too low and slow.

    DRJ (a83b8b)

  91. 91. That’s apparently not he only thing making thm wonder if any of the pilots are lying. The time line they gave for noticing a problem also doesn’t fit the cockpit voice recorder.

    Other developments.

    A third person doed

    It’s definite that one passenger was run over by a fire truck, but they don’t know if she was dead before.

    While fire trucks rushed in, ambulances wer delayed for minutes while they tried to plan how and where the ambulances would go. Passengers caled 911.

    The evaculation of the aircraft was delayed for about 90 seconds. The pilot said not to do it (it is true that passengers can get injured in an emergency evacuation) But then a fire started.

    Sammy Finkelman (a4dbab)

  92. lso the runway was recently shortened by 300 feet which was both good and bad. Good as the approach gave the aircraft more time, bad because this resulted in things being non-operational for months – that glide path or whatever.

    Sammy Finkelman (a4dbab)

  93. re: #92… no, nk, that is just how imbecilic much of the media is, especially in the SF Bay Area.

    Colonel Haiku (1e8f73)

  94. That rady namee Wan Dum Ho, then.

    nk (875f57)

  95. Comment by DRJ (a83b8b) — 7/10/2013 @ 8:39 pm

    DRJ, was that the light in his head from the realization that if he didn’t do something fast, everyone was going to die?

    askeptic (2bb434)

  96. Paging Ron Burgundy…

    Gazzer (9a18d1)

  97. 78.Apparently there were four pilots on board, two to rest while two piloted. That should put to rest the ‘It’s because they were koreans who don’t have relief pilots’ BS.

    Comment by Dustin (dfe6b6) — 7/9/2013 @ 5:18 pm

    Dustin, when he arrived in San Francisco and spoke to reporters the CEO of Asiana said that was standard practice on training flights. Which leaves open the question of whether or not it’s standard practice on all transoceanic flights.

    Earlier, Kevin R.C. O’Brien mentioned this crash:

    (In fact, KAL stuffed a plane in Guam some years ago “flying” a glideslope that was out of service, so as strange as it sounds, it can happen).

    The NTSB report blamed fatique and lack of communication in the cockpit for that crash. There was no relief crew on that flight, just the then standard 3 man crew. The Captain was tired at the end of the flight, made serious errors, and the co-pilot and flight officer didn’t correct him.

    The NTSB report isn’t available online anymore, but the Air Accident Investigations Branch report on the KAL cargo flight that crashed outside of London in 1999 is here:

    http://www.aaib.gov.uk/cms_resources.cfm?file=/3-2003%20HL-7451.pdf

    Pay particular attention to section 2.9, Crew Actions, and section 3, Conclusions. The pilots ignored several serious warnings. They shut off the warning indicators individually three times, and they never mentioned anything about those warnings to each other. The Cockpit Voice Recorder captured the flight engineer calling out several warnings, but neither of the two pilots reacted to that either.

    The report speculates that the copilot might have been too intimidated to warn the pilot of what he was seeing on his instruments, if he understood the significance of what he was seeing. The report notes the captain was much older and more experienced than the copilot, and the captain had criticized the copilot several times prior to takeoff.

    The AAIB mentions the 1997 KAL crash on Guam. No doubt they were aware of the NTSB report which noted the copilot and flight engineer were overly deferential to the captain. The AAIB report states several times that KAL needed to continue to aggressively adapt Crew Resource Management training to Korean culture to correct these ongoing deficiencies.

    Have they been corrected? KAL brought in a retired Delta airlines president to address it, and hired Western pilots as part of their response to that crash (which is where I got my information about Korean airline operations in general).

    All I know, based upon the NTSB’s interim updates, is either none of these pilots were monitoring their airspeed. Or if they were they said nothing about it. Just like the copilot in that doomed KAL cargo flight. At least in this case they said nothing about it until it was too late.

    More on the Asiana Airline crash here:

    http://www.huffingtonpost.com/2013/07/12/asiana-crash-fire-truck_n_3587805.html

    Steve57 (7c82fc)

  98. Via Ace, KTVU is reporting that an intern at the NTSB confirmed those names for them.

    http://minx.cc/?post=341608

    UPDATE [DrewM.]: So much FAIL

    NTSB release: “a summer intern acted outside the scope of his authority when he erroneously confirmed the names of the flight crew…” #ktvu

    — Tim Nelson (@timnelsonPDX) July 13, 2013

    The intern only confirmed the names, he/she didn’t give KTVU the names. Some idiot got those names and called the NTSB. The kid was probably being sarcastic (“yeah, sure that’s their names’) and the idiots at the station ran it.

    There’s video at the link. Naturally the KTVU anchorette calls the summer intern an “NTSB official” in her non-apology, just amplifying the stupidity.

    Steve57 (7c82fc)

  99. The current admin has an unprecedented incompetence level.

    SPQR (768505)

  100. It was definitely conformed on Friday, July 19 that one of the passengers killed was killed on the ground. She was run over not just by one vehicle, but by two.

    Sammy Finkelman (43c045)


Powered by WordPress.

Page loaded in: 0.1246 secs.