TonyWilliams

The crew of a Delta Airlines Boeing 767-300, registration N185DN performing flight DL-60 (dep Oct 18, 2009) from Rio de Janeiro,RJ (Brazil) to Atlanta,GA (USA) with 183 passengers and 12 crew, had declared a medical emergency while on approach to Atlanta. The approach and landing was accelerated, the crew received clearance to land on runway 27R in night conditions but good visibility, however touched down on taxiway M, which runs north of and parallel to runway 27R. The airplane slowed safely.

The FAA reported, that there was no other traffic on the taxiway. An investigation has been initiated, the flight crew has been derostered until completion of the investigation.

The NTSB reported on Oct 21, 2009, that the medical emergency concerned a check airman in the cockpit on the observer's seat, who became ill and had to be relocated to the cabin while still in cruise flight. The runway lights were turned on, approach lights and ILS however were not active. The NTSB are conducting an investigation and have dispatched four investigators.

The FAA released the ATC recordings on Dec 22nd at: http://www.faa.gov/data_research/acc...0_redacted.mp3.

Metars:
KATL 191252Z 08003KT 10SM CLR 06/01 A3032 RMK AO2 SLP274 T00610011
KATL 191152Z 00000KT 10SM CLR 03/00 A3031 RMK AO2 SLP268 T00280000 10039 20022 53016
KATL 191052Z 00000KT 10SM CLR 03/00 A3029 RMK AO2 SLP261 T00330000
KATL 190952Z 00000KT 10SM CLR 03/00 A3028 RMK AO2 SLP257 T00330000
KATL 190852Z 00000KT 10SM CLR 03/00 A3026 RMK AO2 SLP251 T00280000 56000
KATL 190752Z 34004KT 10SM CLR 04/01 A3026 RMK AO2 SLP250 T00390006
KATL 190652Z 33003KT 10SM CLR 03/01 A3026 RMK AO2 SLP251 T00330006

TonyWilliams

The NTSB have released their final report on Nov 15th 2010 concluding the probable cause of the incident was:

The flight crew’s failure to identify the correct landing surface due to fatigue.

Contributing to the cause of the incident were

(1) the flight crew’s decision to accept a late runway change,
(2) the unavailability of the approach light system and the instrument landing system for the runway of intended landing, and
(3) the combination of numerous taxiway signs and intermixing of light technologies on the taxiway.

The NTSB reported, that the captain was receiving special airport qualification operating experience from the check airman, who was commander of the flight.

The check airman was suffering from a gastrointestinal disorder during preflight preparations, but after a brief time away from the cockpit he indicated he was fine and continued with flight preparations. The airplane departed with the captain in the left hand seat, the check airman in the right hand seat and the first officer in the observer's seat. After reaching the top of climb the crew rest times were decided, the check airman was the first to take his break of about 2:50 hours and proceeded to the cabin. At the end of his rest time it was determined that he was ill and unable to continue duties. The crew enlisted the aid of a physician on board.

In consultation with dispatch the captain assumed duties as commander and decided to continue to Atlanta requesting emergency services should meet the aircraft.

Prior to reaching the top of descent the crew briefed the approach to Atlanta's runway 27L. After checking in with Atlanta approach the flight was assigned runway 26R, the crew briefed the landing on runway 26R, but then was re-assigned runway 27L. While passing the outer marker on approach to runway 27L, tower offered runway 27R which the crew accepted. The captain said, while being aligned with runway 27L he maneouvered for the side step and lined up on the "next brightest set of lights" he could see. He saw bright edge and center line lights and thought he had the runway in sight, the airplane however touched down on taxiway M.

At the time of the incident the runway approach lighting had been released into maintenance and was not available. It would have taken about 20 minutes to get the approach lights up. The ILS for runway 27R had also been turned off and would normally have been turned on provided there was sufficient time. However, with an aircraft at or inside the outer marker the time was insufficient.

At the time of the incident the runway edge and center line lighting was operated at low intensity (step 1 of 5, equivalent a current of 2.8 amps). Both edge lights of the taxiway M were operated at high intensity (step 5 of 5, equivalent to a current of 6.6 amps) and the taxiway center line lights were operated a medium intensity (step 2 of 5, equivalent to a current of 3.4 amps). The taxiway lighting at the east end of taxiway M had been replaced with LED type lights in 2009 during an upgrade. Pilot comments had been received that the LED type lights appeared much brighter than the other original incandescent lights but no formal pilot input was collected. The FAA determined that pilots do not like intermixing incandescent and LED lights, but did not publish any standard for intermixing the lights.

Flight tests showed that it was challenging to identify the runway lights of runway 27R as the taxiway lights appeared much brighter. Aligned with runway 27L and side stepping with the same light settings as during the incident flight the runway center line lights were not identifyable and the taxiway lights were much more prominent. While on final to 27R the taxiway lights were still more prominent than the runway edge lights. At 500 feet AGL the runway center line lights were barely visible and it appeared several of them were out. The colour of the taxiway lights became distinguishable at about 500 feet AGL.

The tower could observe the landing lights of aircraft but because of those lights blending in with city lights and the observation angle it was difficult to determine whether the airplane was aligned with runway 27R.

The tower was equipped with airport surface movement equipment (ASDE). The ASDE was not programmed to alert for aircraft landing on a taxiway or departing from a taxiway. Therefore, when the local controller scanned the ASDE display, system safety logic bars suggested, flight DL-60 was properly aligned with 27R. Shortly after DL-60 touched down the ground controller advised the tower that DL-60 had landed on taxiway M. The captain was requested to call the tower after reaching the gate.

In post flight interviews the captain said, that his flight to Rio had arrived on Oct 17th around 09:00L, he took a nap of about 3-4 hours before meeting with the crew for dinner. He returned to his room by about 23:30-00:00L and woke up the following morning at around 09:30L. He stayed in his room, sent e-mails, exercised, tried to take a nap however unsuccessful. In the evening, after being awake for 11 hours, he met with the crew at the airport to begin preflight preparations. He said, he was really looking forward to his rest time during the flight thinking he would get a good amount of rest during that time.

The first officer said in post flight interviews that she had rested well during her layover in Rio taking a nap of about 3-4 hours about 2 hours prior to getting to the airport.

The NTSB annotated that according to the cockpit voice recorder the captain remarked in the approach briefing conducted prior to top of descent, that for this approach the "…highest threat is exhaustion." A member of the cockpit crew yawned at the end of reading a check list and commented regarding lack of sleep. While taxiing to the gate the captain stated he was fatigued when he accepted the last minute runway change and that he had fixated on the VASIs. The CVR indicates by a captain's remark the captain realised they were landing on a taxiway 4 seconds prior to touch down.

TonyWilliams

Why no go-around?

flywithjohn

Let me think, the only time I landed on a taxi way was my first time in a DC-10 sim and I got cocky and asked for lots of crosswind. I don't like to criticize but with 10SM and runway and taxi lights on, how do miss the blue lights for white? At least no one was hurt, what happened to the sick passenger?

DeadHead

As far as the taxi lights being brighter than the runway lights, I have started to see this more and more over the years and always thought it was a bit odd. I guess that rational is for when the airport goes into Cat III operations-Low Vis Taxi ops.

Still it could of gone real bad, I like to at least think that had the crew seen traffic on the taxiway they would have gone around immediately, but either way someone could have joined the taxiway as they were landing.

The fatigue thing definitely can catch up to an experienced crew though. I think ATC in ATL could be better about giving runway changes, sometimes I feel as though it's not even worth briefing until they officially assign you a runway.


These guys were lucky for two reasons,
1. Everyone walked away ok and there was no damage.
2. The Laptop-MSP incident seemed to take the spotlight off them.

rickair7777

RWY lights are DEFINITELY being set to lower power levels on a routine basis compared to three years ago. Cost saving I assume.

But why no go-around? That is perplexing...even at T/D it seems safer to abort than risk who-knows what collision hazards downrange (assuming advancing the levers will cancel autobrakes, lift-dumpers, etc).

dojetdriver

You mean sick LCA?

Couldn't agree more. Prepare to be criticized/lambasted by those "in the know".

flywithjohn

Yeah, I know I made an error. They happen

dojetdriver

Indeed they do, like, oh, I don't know. Landing on a taxi way

tortue

I wonder if its possible to reconfigure a sim to have it "move" you offset of a runway a few hundred feet above while you're about to touchdown to an adjacent taxiway to see the reaction. The idea being to simulate the psychological experience of where you totally believed you were lined up and over the threshold but at the last few seconds were not to see if you'd go around or just plant it where it was. Would be interesting to try it at 50 above, 100 above, etc.