Friends have been asking me about the short-of-the-runway crash of an Airbus A320 at Halifax this morning. This article says that the pilots were trying to land on Runway 5 (oriented northeast; reasonably appropriate for the wind at the time). fltplan.com shows that the most likely approach to Runway 5 for an airliner is a “non-precision” approach without vertical guidance. These tend to be trickier to fly than a standard ILS. (I have written about the dangers of these approaches in an article on instrument flying.) Most U.S. airlines and, I think, Canadian charter and scheduled operators, are required to try to turn non-precision approaches into quasi-precision ones. The pilots calculate a rate of descent that should keep the plane on a reasonably stabilized glidepath toward the runway. If they don’t see the runway when 100′ above the minimum descent altitude, they add power and go somewhere else. When I flew regional jets for Comair/Delta we called this “CANPA” (constant angle non-precision approach) as distinct from the “dive and drive” technique that the FAA traditionally promulgates. This article by a Gulfstream pilot does a great job of explaining the alternatives.
It is atypical to have to do a non-precision approach in the airline world. Most big airports have instrument landing systems (ground-based radio beacons providing left-right and up-down guidance) on the frequently used runways.
It is atypical to crash an airliner. But crashes are more common following non-precision approaches (a classic one). That it was dark, late, windy, and snowy probably did not help.
[The airport has a precision “GPS/LPV” approach to the same runway. These can be flown very straightforwardly in a light aircraft by upgrading the $10,000 GPS. But the forest of regulation around the equipment that goes into an airliner could make an upgrade to handle these kinds of approaches cost closer to $1 million per plate. The professional pilots in this forum talking about the accident suggest that Air Canada does not have this kind of capability, which has been standard in little airplanes for about 6 years.]
Related:
I thought that the terrain warning systems were supposed to prevent this type of accident. Why didn’t it, or is this type of accident so close to the start of the runway beyond their capability?
Izzie: That’s a great question. I don’t know. Certainly the EGPWS systems have a database of runways and are supposed to suppress warnings when you’re lined up and close to a runway while also close to the ground. If they could tell the difference between 2000′ short of the center of the touchdown zone (this Air Canada flight) and spot-on then you wouldn’t need any extra equipment to fly an LPV approach (since you could use the GPS built into the EGPWS!).
And then there is also the radar altimeter. Nova Scotia is fairly flat. So they should have gotten a clear read-out of height above the ground. Perhaps they just mistakenly thought that they had seen the runway and were intentionally landing on what turned out to be unpaved ground instead. (LPV approach would still have likely prevented this because airline practice is to continue to use the flight director for reference almost all the way down to the pavement.)
The $100 GPS I have in my car can tell my position to within a few feet, so I can’t believe that the GPS that is in the terrain warning system can’t pinpoint an aircraft’s position to within 2,000 ft. There must be lots of places on earth where you would be well above the terrain at position X but 2,000 feet away you would be in big trouble.
I also thought that the EGPWS Mode 5 is supposed to warn you if you are below your glide slope? (My knowledge of this begins and ends with the wikipedia article).
I have another question. Why was the plane NOT flying ILS in the middle of a blinding snowstorm in the dark of the night? Isn’t this exactly the kind of situation for which ILS is designed? Apparently Runway 5 does not have ILS but 23 (same runway from the other direction) and 14 do. Did the winds not permit these other approaches?
The passengers were very lucky that the spot where they crashed (ahem, made a “hard landing” according to Air Canada) was relatively flat and that they were able to belly skid onto the runway. One or two slightly more solid obstacles could easily had send the plane tumbling wing over wing or started a fire or sliced thru the cabin but in this case there appear to have been no serious injuries even though they did considerable damage to the airplane.
Airport |Authorities and the Federal government are to blame they have spent billions building monuments to them selves but spent little on upgrading the runways, taxiways and navigation equipment. This runway should of had an ILS not a Non P approach .This is happening all over the country lets only put in one ILS and then put in a RNAV Approach because its cheaper! For an airport like Halifax that gets such crap weather all runways should of had ILSs!
It strikes me that ILS is a sunset technology and that future efforts should go into things like LAAS/GBAS. But even this is only planned for the largest airports.
But as Phil points out, the problem is that the cost structure of the aircraft industry (partly as a result of FAA requirements and partly to account for the potential massive liability costs in the event of an accident) means that such systems cost the airlines hundreds of times more than they should.
Existing (unsafe) equipment and procedures are “grandfathered” and free but anything new that would add safety has to be verified to an almost impossibly high standard, so it doesn’t get added at all. Rationally, if we could add a system that prevented 10 crashes of the Air Canada type but might be responsible for 1 crash of a different type, we should implement such a system – society would be much better off achieving a net reduction of 9 crashes. But under the current system, we can’t add such a system unless it has near zero risk of causing any crashes of its own (achievable at a very high cost) so instead we live with the current unsafe system.