Friends have been asking me how it is possible that the Asiana 777 landed short of the runway at SFO.
It turns out that the instrument landing system (ILS) glide slope was out of service. So the pilots were likely conducting a “visual approach”, i.e., looking out the window to see if the airplane was properly positioned to land. There are synthetic glide slopes available from WAAS-enhanced GPS receivers and there is a RNAV 28L approach at SFO that provides LPV minimums for such units. These cost about $500 to install in an experimental (uncertified) airplane, but regulation adds $10,000 to that cost when the gear goes into a certified airplane such as a four-seat Cirrus SR20. Touching even one screw on an airliner, after it leaves the factory, costs closer to $1 million. So it seems unlikely that Asiana had retrofitted their B777 with the latest WAAS GPS gear.
How hard it is to fly a visual approach? That’s how all approaches are flown in light piston-powered airplanes during training. Things move both a little faster with jets and a little slower. The speeds are faster, e.g., 120-145 knots on final approach instead of 60-70. What is slower is the response of the aircraft to power adjustments. Some of this is due to simple inertia. The airliner weighs a lot more. Some of this is due to the fact that jet engines, once “spooled down”, can’t provide instant power.
Here’s a story from my second month of flying regional jets for Comair, a Delta Airlines subsidiary. I am leaving the original language but adding notes in brackets because this was written for my pilot friends.
Today was a beautiful clear and calm day for flying. I took the CRJ [Canadair Regional Jet, a 50-passenger 53,000 lb. airliner] from RDU up to LaGuardia. The ATIS [canned weather broadcast, updated hourly] said that we could expect the ILS 22 at LGA. Our clearance was to descend and maintain 4000′, head for the Verrazano Bridge, then fly up the Hudson River. We were high and close to the airport, restricted to a minimum speed of 180 knots, when New York Approach cleared us for the “visual 22”. These are the toughest maneuvers for newbies because one has much less time to get stabilized than with a full ILS procedure and it is almost impossible to use the autopilot, which won’t intercept the radio beams at extreme angles.
The airline encourages us to fly 5 knots faster than “Vref” [around 145 knots in a CRJ with all seats filled] during training, in order to have a margin for error in case of wind gusts or incompetence. When a runway is short, however, the extra energy is difficult to dissipate and tends to result in significant float. The runway at LGA is 7000′. A test pilot demonstrated the ability to get the airplane stopped in about 3000′, but it wouldn’t necessarily have been consistent and it certainly would not have been comfortable for passengers. The minimum runway length that people regularly use for the CRJ is 6000′.
My approach was slightly fast and slightly high. I was reluctant to adjust the thrust in the last 300′ or so because we had been fairly stable on glide slope and on airspeed. In previous approaches I had tended to overcontrol. I started the round-out at the 1000′ markers, and touched down about 2500′ down the runway, applying moderate brakes and thrust reversers and turning off at Charlie, having chewed up almost 5000′ of runway (out of 7000′). Passengers complimented me on the smoothness of the landing, saying “It is usually rough here at LaGuardia” (of course it is usually rough because pilots more skilled than I are trying to get the plane on the ground, have the spoilers deploy, and stop the plane before it runs into a swamp).
Captain Mark said “That landing sucked. I’ve got 13,000 hours in this airplane. I’m going to show you how it is done on the next leg.” He then asked “Are you flying the airplane or is the airplane flying you?” When I asked for specific tips Captain Mark replied that I was an instructor and ought to be able to figure out what I was doing wrong. He grudgingly confirmed that I had left too much thrust in for too long.
We approached Charlotte, North Carolina in near perfect conditions. The weather was smooth and, due to some clouds at 2500′, we were given vectors for the full ILS 23 approach. [The controllers vectored us so that we were lined up with the runway approximately 10 miles from touchdown.] Captain Mark let the autopilot do most of the work, concentrating on getting the thrust exactly right for a stabilized approach. At about 200′ above the ground, Captain Mark disconnected the autopilot and transitioned to hand-flying. Somehow he ended up a little fast and also flared a bit too high. The CRJ is a very efficient glider and spooled down jet engines don’t supply the kind of drag that props would. The CRJ entered a shockingly efficient glide in ground effect at 10-15′ above the runway. We weren’t descending. We weren’t slowing down. The 7500′ runway was slipping away beneath us.
It is unclear how one would fix a situation like this. [In a piston airplane the best and easiest fix is to add power, retract the flaps, and climb away from the runway in order to try again; this can’t be done in an airliner due to the long spool-up time of the engines.] In a piston airplane you’d add a touch of power and pull back for a slower and less efficient airspeed. The airplane would sink due to loss of efficiency and the power would slow the vertical speed. In the jet, once the thrust levers are back it takes 3 or more seconds to get any significant power from the engine. Nosing the airplane forward would result in hitting the nose gear, which isn’t any better on a jet than on a piston four-seater.
After we had sailed over approximately half the runway, the airplane finally started to settle towards the surface. We touched with about 3000′ remaining [i.e., 4500′ down the runway, 2000′ more float than I experienced at LGA]. Captain Mark slammed hard on the brakes, to the point where the passengers probably would have said “ouch!”, and applied full reverse thrust. Tower called and asked “Are you going to be able to make Foxtrot?” This was the second-to-last taxiway and only about 500′ from the end of the runway. In fact, we did make Foxtrot, but only barely. We used just about a full 7000′ of runway. [I.e., at LGA we would have been nose-to-nose with the boats.]
What did I learn from watching a very capable guy with 13,000 hours of CRJ experience? That landing a CRJ consistently requires more than 13,000 hours of experience…
————— end of story from 2008
Additional background: The CRJ is an adapted business jet and, lacking leading edge devices or “slats”, lands much faster than a standard airliner such as a Boeing 737. In addition to the challenge of speed the pilot must, in the last 40-50 vertical feet of the flight, pull the airplane from its 3-degrees nose-down attitude to a standard nose-up landing attitude of about +9 degrees. (This lead one FlightSafety instructor to refer to every landing in a CRJ as a “controlled crash”. The procedure is different enough than in a Boeing or Airbus that the plane comes with a special briefing card for jump-seating pilots of conventional airliners so that they don’t start screaming in the last 30 seconds of the flight.)
Follow-up: Folks to whom I emailed the story asked me how the rest of the trip with this pilot went. My reply: “Captain Mark found fault with everything that I did for the next three days. I hadn’t ironed my shirts properly. I left my bags next to the airstair door while doing a preflight inspection, thinking that it would give the captain and flight attendant more space and freedom to put away their bags. Captain Mark admonished me “You have to go up into the airplane, put your stuff away, and then go back outside to do the preflight.”
Punchline: About six months later, US Airways 1549 was flying the same route (LGA to CLT) and went down in the Hudson River. Asked my opinion about the heroes Captain Sully and Jeffrey Skiles I said “Hey, Captain Mark and I took 50 passengers from LGA to CLT. We got them to the gate at Charlotte, on time, warm and dry, and nobody called us heroes.”
The PAPI was also out so the crew did not have vertical guidance. There is some reasonably plausible explanation for this tragedy at PPrune http://www.pprune.org/rumours-news/518568-asiana-flight-crash-san-francisco-15.html
Being high, fast and then over correcting in a jet without vertical guidance while fatigued from a Transpacific flight is a possible scenario. There are some FLC / auto throttle issues that may be a factor as well.
Well It seems to be a “fat ass sensor” malfunction in other words the speed went down while the pilot was looking outside. the airplane started to react very slow the pilot did not realize that becuse he has been flying this giant with the eletronic equipment while it would be better to fly using his fat ass sense linked to the throttle lever in his hands.
These giant birds are better flown by pilot`s body motion senses.
Listening to the audio of the communications between the air traffic controllers and the pilots it is outrageous that the system is so old fashioned. Why is a lot of this not automated?
I guess I know why: it is run by the government.
I live in the Frisco bay area (I use this to annoy the snotty locals) and I have learned to avoid SFO. At the least provocation (wind above 20 knots, slight cloud cover) there are always long delays. AFAIK the main reason is that the 2 main runways are too close for all but the most ideal conditions since the enviros will not let the airport fill in a few hundred feet of the bay to separate them. They then switch to one runway, which of course clogs everything up.
Now I learn that another reason is that the airport operations continue without major safety equipment like this “ILS glide slope” system. Heads should roll in the airport administration but of course will not.
(I have no experience flying aircraft.)
It looks like there’s a lot of armchair crash investigation happening, mostly using FlightAware’s record of the plane being 85kts at 200ft. Is it possible to work out how abnormal it is to significantly botch a landing like this using something like FlightAware’s data set for other flights? I’d be curious to hear how often visual landing of jetliners happens at SFO, and how often the landing ends up being this far “out of spec” (without fatal consequences until now, obviously).
Bob,
“Now I learn that another reason is that the airport operations continue without major safety equipment like this [‘]ILS glide slope[‘] system.”
The ILS glide slope is indeed a piece of safety equipment, but it is completely unnecessary for operation in visual conditions, though it can be used by the pilots if they choose to, in visual conditions. It is only important, and indeed vital, when operating under low visibility conditions, such as fog, heavy precipitation, or low clouds. In those conditions, when the pilots cannot see the runway in front of them, they use the radio signal from the ILS Localizer and Glide Slope to guide the plane down to the runway.
Furthermore, the FAA had sent out a NOTAM (Notice to Airmen) that it would be out of service at the time. The Pilot in Command of an aircraft always is the ultimate authority for the operation of an aircraft. If the pilots didn’t feel safe landing without an operating glide slope, it would have been their prerogative, and responsibility, to determine and request an alternate plan. For example, they could have requested that the tower assign them a runway that had an operating glide slope. That may have meant a delay while ATC shifted traffic to accommodate their request, but I guarantee you they would have, had the pilots requested it.
Tens of thousands of flights safely fly and land every day without glide slopes. Indeed, only a small minority of runways even have operating Instrument Landing Systems, and yet student pilots with only 10 or 20 hours of flying experience manage to land safely on them hundreds, maybe even thousands, of times every single day.
The PAPI’s were operational at the time of the accident.
After reading your Dec 2009 post on foreign airline safety (http://philip.greenspun.com/flying/foreign-airline-safety) I’m curious about how many hours as pilot-in-command the Asiana pilot had. Does he have a flight instructor rating? What about the other categories of hours in that article? In short, how relevant is the “10000 hours” quoted in the media?
The automatic airspeed control was dialed in but not enabled, 1 of the thousands of things they have to do to land. That combined with the lack of ILS, the lack of a buffer zone in front of the runway, & a speeding firetruck hitting 1 of the victims would have conspired to make it a bad day. There will now be buffer zones extending into the water, eliminating this problem until someone finds another problem.
> I’m curious about how many hours as pilot-in-command the Asiana pilot
“The junior pilot of the Asiana jet, 46-year-old Lee Kang Kuk, was at the controls and was in training on the 777 with just 43 hours of experience on that model, and 9,793 flight hours, according to the airline.” — http://www.bloomberg.com/news/2013-07-10/asiana-crash-shows-lessons-of-pilots-trumping-technology.html
I know that the media were reporting 10,000 flight hours. But I’m asking how many hours as *pilot-in-command*.
Bradley: I think that the flying (in training) pilot had plenty of pilot-in-command hours but he had no experience as a flight instructor, according to the article that I saw. As is typical with foreign carriers he had been recruited directly into the airline’s training program. So all of his experience would have been pushing buttons in big jets while sitting next to competent big jet pilots.