Bureaucrats at Work

This is your Captain speaking…

One of the things that has always bugged me from day one as a pilot has been the way the aviation bureaucracy reacts to incidents and accidents. Time and time again some idiot screws something up, and the powers that be react by making whatever he did illegal. Now the National Transportation Safety Board is doing the same thing in a reaction to the Southwest accident at Midway airport in Chicago that tragically killed a 6-yr old boy.

Southwest Airlines Flight 1248 slid off the end of runway 31 at Midway in a snowstorm on a short, outdated, (cheap) runway with high (obsolete)(cheap) minimums. They landed with a 7 or 8 knot tailwind and more of a crosswind than that. They used a computer in the cockpit that figured out how much runway they needed and confirmed their figures with their dispatch function, which shares with the Captain responsibility to operate the jet safely. In the actual fact of the matter makes them just a place to bounce ideas off of and another person to look up any regs or performance factors that the Captain might need to help him make the final decision about the conduct of the flight. (The Captain has the final say because it is his life on the line, but the FARs split the authority and the responsibility.) When they used the computer they put in the use of the engine reversers, a logical decision because the reversers shorten the ground run, and make it legal to use a shorter runway.

The NTSB has just put out a request to the FAA to not allow the use of reversers in the calculation of determining whether or not you can use a certain runway. This has always been the case in preflight planning, but like Michael Yon says, “A plan is just a guess in a fancy dress.” Once you are in the air, a pilot can use the increased performance of the engines in reverse to see if s/he can legally use a runway. It is this choice that the NTSB wants to take away from the pilots.

Normally, this is not a problem, especially if you use airports that have modern grooved runways that do not have any ice on them. Unfortunately, that may not be the airport and runway set you are on the way to as you approach your destination in the winter. As daylight ends, and your fuel spins out the tailpipe, you are likely to take any runway you can get to that you can stop on. If it happens to be icy, the only way you can stop may be the reversers, because the brakes only stop the wheel. If the tire slides and the antiskid brakes release, the only thing the pilots have going for them for sure are the reversers. Without being able to use the reversers in the runway-required calculation, the runway required can easily exceed runway available on ice or packed snow. Not a really great situation, but one which happens every now and then every winter. One pilot making a number of mistakes should not cause a rule change which impacts the whole industry, because the runway required is not greater then the runway available if you can use your reversers in the calculation and then do everything right when you land.

Unfortunately, the Captain on Southwest Flight 1248 did not do everything right. They were going to Midway, a secondary airport in the area, which is typical of Southwest’s operation. The wind was out of the southeast, but Midway has only one direction of landing with even halfway low minimums. The weather was below minimums for all the other landing directions except for the northwest, so they had to land with a tail wind of about 7 or 8 knots. They consulted with dispatch, using reversers to determine if they could legally use runway 31 with the tail wind. The weather sequence did not have the magic phrase of “Braking advisories in effect” so they did not have to take into account the slick runway in their calculations, but I do not know if they did so anyway. Certainly the runway is plenty long enough if it is clear and dry, but they had to use the reversers to get legality, so they took into account the slick runway. (If braking action is less than good, [fair, poor, or nil], the weather sequence should say “Braking advisories in effect.” This alerts the pilots to a less than ideal runway surface. Every landing jet then should advise the tower of the braking condition they felt on their landing roll, and the tower should relay that to the next jet as they change over to tower frequency. This gets the most current information possible to each pilot as they approach their landing.)

According to the early reports, Southwest Flight 1248 landed a little long. The NTSB has had no comment on the initial touchdown point yet. They landed a little hot at 143 knots, but accelerated on the runway to 152 knots before the reversers deployed. Less than 20 seconds later(17) they were out of the airport boundaries, bouncing over a busy road toward tragedy. It took the reversers 18 seconds to deploy, which is why the NTSB wants to prohibit the use of the reversers to determine runway suitability. The Captain reported that he could not get the reverse levers to raise. After about 18 seconds, with the jet accelerating and the end of the runway rapidly approaching, the First Officer deployed the reversers without a problem. Both pilots then applied full brakes but it was too late to stop the jet before it departed the runway.

I disagree with the NTSB request. In order to deploy the reversers, the pilot pulls the throttles to idle, lifts his hand off the throttles to smaller levers in front of the throttle levers and pulls them up. The lifting of these reverse levers causes a movable panel on the engine body to slide aft, closing some blocker panels that reverse the flow of fan air to the front, resulting in negative thrust. The higher you raise the reverse levers, the more power you put into the engine and the more reverse thrust you get. The reverse thrust levers will not raise if the main throttle levers are not in the idle position, and the reversers will not deploy. Reading the sequence of events leaves me with no choice but to conclude that the Captain did not have the throttles fully in the idle position, precluding the reversers from deploying. This may also have had the effect of not allowing the autobrakes to engage. (I never flew the 737, especially one this modern, so I do not know how many interfaces the autobrakes may have.) In any case, the jet did not stop in time. (As late as June, 2000, Southwest Airlines had deactivated the autobrake systems in all their jets. Autobrakes stop the jet much faster than manual braking, but since some of their older 737 did not have autobrakes installed Southwest deactivated the autobrakes in all their jets, according to the accident report on the Southwest jet that slid off the runway at Burbank.)

In all accidents, there is a chain of events and circumstances all of which have to go wrong in order for the accident to happen. There is a great number of them here. The weather had a very narrow band of heavy snow, but the place with the heaviest concentration of snowfall in the whole storm was the 10 inches that fell on the Midway airport. O’Hare, just a few miles north, had but a few inches. Midway’s approach aids are old, but even worse, not all the main runways have the lower minimums of an ILS. If Southwest Flight 1248 had been able to land going into the wind, with a 7 or 8 knot headwind, the NTSB says the landing runout would have been 1000′ less. The approach aids going in that direction have higher minimums, higher that the weather that night. Still, if they had landed nearer the end of the runway they had, gotten the reversers deployed in the normal second or 2 and been able to get the brakes, placed on auto/max, working earlier in the landing runout, this flight would not have ended in tragedy. I’m sure there is no one who feels worse about this whole thing than the Captain. I know that I always felt that the one thing I wanted to do was to not cause harm to others by any of my flying, and I’m sure he feels the same way. He was just unable to break the chain of events he found himself in, the chain of events that caused the only fatality in the industry in two years.

The NTSB wants to break this chain by not allowing the use of engine reversing in calculating runway required for landing. As far as I know, this is the only accident in history that would have been prevented by this proposed rule. Why should we change something pilots may need during flight over a single accident in the total history of jet flight, some 50 years? One time in 50 years? There was just too much pilot error in this incident to take from all our pilots something that is used on an occasional but consistent basis to good effect. The FAA should not implement this recommendation.