Comair CRJ crash in Kentucky

[Astra driver]

Originally Posted by Shore Guy

One of the many ironies of operating a “sophisticated” transport aircraft is that your rental car will probably be better equipped for surface guidance (GPS driven map) than the aircraft you are in.
A piece of paper (10-9) and a wet compass for guidance…..that is (along with signage) the normal way to find your way around some of the busiest and most complex airports in the world.
Surface maps are slowly making their way into cockpits via design (A-380, B 787) or as afterthoughts (EFB’s). They are long overdue.
Honeywell has a system that may have saved the day:
http://www.honeywell.com/sites/aero/...4E407D99FE.htm
And others are working on similar systems.

I have the pleasure of occasionally flying a GIIB that is equiped with this honeywell system, and I can attest that it is a great aid to have in the cockpit, while there is no guarantee that it could have prevented this accident, it would have provided 4 additional cues that would have told them which runway they were on;

1.) A complete Jeppesen airport diagram presented with real time aircraft location presented in moving map format.

2.) An aural warning when approaching rwy;
"Approaching runway 26"

3.) An aural anouncement when on rwy;
"On runway 26"

4.) A runway distance remaining anouncement;
"three thousand five hundred feet remaining"

An excelent and easy system to use. One wonders, given that the FAA is putting priority on reducing runway incursions, that this accident may become the impetus to require all air carriers to install such a system.

[gwillie]

Witness to crash describes what he saw

http://www.kentucky.com/mld/kentucky/15382605.htm

[DanAir1-11]

Originally Posted by jondc9

for ATC to not visually scan the runway as the plane was cleared for takeoff would be a mistake.

Fine, except that in the half light of morning, or earlier, the controller could well have done (and probably did) a visual, and would have seen a set of strobes and lights at the intersection, as he/she would have fully expected ('if' this was an intersection t/o, which has still not been established) What he/she would not have been able to establish, would have been orientation whilst the a/c was at the intersect. There can only be approx 40 degrees difference in heading orientation of the a/c at the intersect between the two rws. Once rolling, it would take a few seconds for the controller to register that the a/c was on the wrong rwy. Again, the light and brg difference early in the roll would make this difficult to detect until sadly it would have been too late. So let us not be too hasty in speculating the twr has erred here.

Regards
DA1-11

[jondc9]

DA1-11 and ATC watcher.

I've spent time in towers around the country...they are quite tall with huge windows as you know.

If parts of the airport are not visible to the tower, a note is made on the charts to let pilots know.

If there was ONE controller at KLEX and he was being forced to do too many things to NOT LOOK OUT THE HUGE WINDOW, then the system needs fixing.

AND one plane with all landing lights on should have been quite distinct especially if taking off on a runway WITH NO LIGHTS.

j

[david foster]

gps?

GA pilot here. How standard are moving-map GPSes in airline cockpits? Is it safe to assume that this plane probably had one, or not?

Do these GPSes have a zoom capability that would show the airport layout, or would a separate device, like the Honeywell system mentioned above, be required for that?

[DanAir1-11]

Rwy 8 / 26 Info

FYI Chaps

Runway 8/26
Dimensions: 3500 x 75 ft. / 1067 x 23 m
Surface: asphalt/concrete, in poor condition
CONC IS SEVERELY CRACKED.
Weight bearing capacity: Single wheel: 12500 lbs

Runway edge lights: medium intensity
MIRL RY 08/26 OTS INDEFLY.
RUNWAY 8 RUNWAY 26
Latitude: 38-02.339143N 38-02.433593N
Longitude: 084-36.574985W 084-35.855642W
Elevation: 969.5 ft. 972.6 ft.
Gradient: 0.3 0.3
Traffic pattern: left left
Runway heading: 085 magnetic, 081 true 265 magnetic, 261 true
Markings: basic, in fair condition basic, in fair condition
Runway end identifier lights: yes
REIL OTS INDEFLY.
Touchdown point: yes, no lights yes, no lights
Obstructions: 25 ft. tree, 400 ft. from runway, 150 ft. right of centerline, 8:1 slope to clear 17 ft. pole, 552 ft. from runway, 39 ft. left of centerline, 20:1 slope to clear

source - http://www.airnav.com/airport/KLEX

Regards

DA1-11

[Airbubba]

Some information from the 10 pm NTSB briefing in LEX:

Aircraft takeoff weight was 49,087 lbs. according to the load manifest.

The manufacturer calculates that 3539 feet are required to rotate at this weight.

Runway 26 lights have been inop for years and cannot be turned on by the tower.

Runway 22 centerline lights were OTS, the other lights on Runway 22 were reported on by a witness.

The FO was flying pilot.

Takeoff roll was 29 seconds.

A comment was made during the takeoff that the runway lights were out.

137 knots was the max airspeed attained.

The captain called '100 knots', 'rotate'.

No calls from ATC during the takeoff roll.

A team is out tonight taxiing a Comair RJ to see what the perpective was for the crew in the darkness. People are also in the tower to check visibilty of the taxiing plane.

Engines were turning, spoilers and T/R's were stowed, no evidence of braking.

The aircraft rolled off runway 26, left a trail of all three landing gear on the grass, went through the boundary fence, went up a berm and became airborne, cleared a barbed wire fence and hit a tree line. The wreckage was about 900 feet past the tree line.

The FO is still in critical condition and has not been interviewed.

The FDR had 51 hours of data, 33 flights.

Many of the investigators are going to CVG on Tuesday to interview management and pilots at Comair and to retrieve training and personnel records.

[Cross-eyed]

Taxiing to RW22 RW26 at KLEX

Photos and diagram from kentucky.com Lexington Herald-Leader

Also from the Herald-Leader:

[Meanwhile, a former Delta Airline pilot said yesterday that flying out of Blue Grass Airport can be confusing and that straying onto the wrong runway would have been “an easier mistake than people generally would think.”

“It looks like a no brainer ... but it is possible to be kind of confused,” Russ Whitney said.

Among other things, Whitney said, the main Runway 22 has a crown or hump, so that pilots taking off cannot initially see the southern two-thirds of the runway. As a result, Whitney said, Runway 22 and the shorter Runway 26 can look as though they are the same length.

“I’ve taxied out there and gotten kind of confused, and had to make absolutely sure that I was on the right runway,” he said. “I have taxied out ... and said to the co-pilot, ‘Is this the right one?’”]

[DesiPilot]

I worked as an FO over a year flying CRJ's. I can tell you from my personal experience that an FO is really really busy while taxiing out. Not only we have to run the checklist but have to calculate TO performance, V speeds, feed the info in the system.

We had ACARS on our planes and everything was done through ACARS and still we were very busy. I understand COMAIR does not have ACARS and the FO has to manually calculate all the numbers while taxiing.

So, the FO may not be looking outside where the aircraft was taxiing. Also if they were flexing, it may increase that 3539' length.

A very sad day in aviation.

[Scurvy.D.Dog]

jondc9
DA1-11 and ATC watcher.
.
I've spent time in towers around the country...they are quite tall with huge windows as you know.
.
If parts of the airport are not visible to the tower, a note is made on the charts to let pilots know.

.. what about the haze?

If there was ONE controller at KLEX and he was being forced to do too many things to NOT LOOK OUT THE HUGE WINDOW, then the system needs fixing.

…. we do not know that was the case!

AND one plane with all landing lights on should have been quite distinct especially if taking off on a runway WITH NO LIGHTS.

… what is the position of the tower relative to the runways?? …. could they see the runways that morning?? ….. any (P)SMR display at Kentucky? …. squat switches might have precluded transponder transmission?!?!
.
… either way …. There were lots of factors in this one ….. lets not unfairly tar any individuals!!!

[OKFINE]

Yes indeed a sad confluence of events. Not to appear sarcastic but I assume there were NOTAMS regards the runways during the resurfacing etc. The crew flew into the airport so had a view of the field with its idiosyncracies. Sometimes merde s'arrive. I know a chap who started an A-320 with the thrust levers NOT at idle. This despite a ramp check, a before start check and an ecam warning "thrust lever move". I guess some times we see what we want to see. God rest their souls.

[archae86]

Originally Posted by Airbubba

137 knots was the max airspeed attained.
The captain called '100 knots', 'rotate'.

If these calls were in close proximity, and if 100 knots is appreciably below standard rotation speed for the conditions, may this suggest captain at this point realized a problem, and thought the chances better at trying to fly out than to do a very high speed over-run?

Anyone care to post an estimate of rotation speed for the given conditions?

[flyguykorea]

Originally Posted by er340790

Aircraft takeoff weight was 49,087 lbs. according to the load manifest. The manufacturer calculates that 3539 feet are required to rotate at this weight.
Runway 8/26 Dimensions: 3500 x 75 ft.
Thirty-nine feet. Must have been desperately close.

A good argument for keeping as much runway available as you can when lining up. I remember from an old Boeing instructional video "runway behind you, is runway you can't use during a RTO". While a RTO doesn't apply in this situation, it clearly shows that every little bit may have counted. I guess at this stage we don't know if they had a rolling or standing takeoff, but I would presume that they would have lost 100ft or so in the line-up.

[Rongotai]

Looking at the maps and photos posted by Cross-Eye I am struck by the possibility of confirmation bias factors in this accident. By this I mean the cognitive habit of seeing what you expect to see and discounting contrary information if it is not jarringly contradictory. This appears to have been an important factor in the Erebus accident, with the crew 'seeing' McMurdo Sound because that is what they expected to see.
If so then it isn't necessary to postulate fatigue, but the overall early light environment does come heavily into play, together with the change of the taxiway configuration.
One of the problems with confirmation bias is that it can happen to anyone at any time - even the very best in optimal conditions - and that it is the one cognitive phenomenon that can easily be magnified in a well-functioning CRM environment.
If my speculation is correct I would then want to suggest that when reconfiguring airport layouts there be positive identification of the risk of producing taxiway and runway patterns that are sufficiently similar that they might induce confirmation bias errors. The pre-existing configuration shown would make such an error difficult (not impossible) to make, but the new layout increases the risk markedly.

[Ignition Override]

JonDC9:
Many of these 'contributors' might have little aviation experience, or they (pilots do this?) assume that that a Tower Controller (supervising at least 3 frequencies + some telephones?) can/could always notice and keep track every aircraft location and direction of movement, as you already tried to clarify. Maybe the CRJ in Lexington had its left or right landing light deferred per MEL on the flight release. Even a nosegear taxilight could be inop., in order to keep Dispatch reliabilty high.

At larger airports, very experienced ground controllers (ATC) easily lose track of which aircraft is holding at a taxiway intersection or a de-icing 'pad' due to wx or security glitch delays, as happened to us tonight at a large midwestern airport.

Low visibility happens very quickly even after a very hot, dry summer.
Right now, seen from the ground by these hotels, fog appears to be forming over Lansing, MI. It was only 18*C when we arrived earlier, with the dewpoint at 17*, altimeter 29.84: the first cool temp I've seen since a vacation in Normandy France in May. What a relief from down south .

[Brit50483]

The same turn angle onto the 'active'

Looking at the maps posted by Cross-eyed in post #158, it struck me that the left turn onto R26 from taxiway Alpha (about 140 degrees) looks almost eaxactly the same as the left turn they used to make from the now disused taxiway onto R22.

By comparison, the turn now required off Alpha to get to R22 is about 90 degrees, markedly different.

Maybe they made the 'same' turn they always made and saw a runway ahead of them ...

[Metro man]

Since the SIA accident at Taipei, Singapore regs require the departure runway to be identified. We do this twice, on taxi check departure runway/SID in the box, and on line up "Departure runway XX identified" both pilots. Either check the signs at the holding point, the numbers on the runway or the heading on the nav display(remember if parallel runways available)

[Sensible Garage]

computed data

shoot me if I'm wrong but can't be far off

Field elev 979'
press assumed 1015/2992
tep assumed 15
wind zero assumed
anti ice off
bleed zero (on APU)
APR armed
dry
20 deg flaps
TOGW 49086
take off distance 5629'
V1 131
VR 135
V2 140
VFTO 181
%N1 92.8

I've no idea how they come up with 3539 feet...
100 knots for VR does not sound possible to me in a CRJ200

[A7700]

Originally Posted by jondc9

for ATC to not visually scan the runway as the plane was cleared for takeoff would be a mistake.
I recall being cleared for takeoff on long runway at KMRY, beautiful day, clear and a million, only plane on airport.
I asked controller if he was sure I was cleared for takeoff...he got quite irate at my question.
the problem was a fuel truck on the runway about 4000 feet from threshold.
HE hadn't even LOOKED at the runway and cleared me for takeoff. KMRY is noted for fog, lucky that day it was clear enough for me to see.
ATC is responsible for seperating all IFR aircraft, and not to scan the runway ( and see, by extension correct runway/aircraft configuration) would be WRONG>

ICAO DOC4444

7.1.1.2 Aerodrome controllers shall maintain a
continuous watch on all flight operations on and in the vicinity
of an aerodrome....Watch shall be maintained by visual
observation

7.10.1 Control of aerodrome surface traffic
in conditions of low visibility
Note.— These procedures apply whenever conditions are
such that all or part of the manoeuvring area cannot be visually monitored from the control tower.

7.10.1.1 When there is a requirement for traffic to
operate on the manoeuvring area in conditions of visibility
which prevent the aerodrome control tower from applying
visual.......then 7.10.1.1.1 and 2 = PROCEDURAL CONTROL

Note.— The Manual of Surface Movement Guidance and
Control Systems (SMGCS) (Doc 9476) provides guidance on
surface movement guidance and control components and
procedures for low visibility operations


At night, with in addition some rain showers and haze, the controller can only count on the pilot's reports. Even with an SMR you can only assume that a runway is vacated when the pilot declared he has done it.

[fepate]

Originally Posted by Sensible Garage

...
take off distance 5629'
...
I've no idea how they come up with 3539 feet

Your figure is the accelerate/stop/go balanced field length, no? The pertinent figure here is the actual distance to liftoff with all engines operating.

100 knots for VR does not sound possible to me in a CRJ200

After going off-road, through a fence and approaching a clump of trees it may be worth a try, though.