Airline Crash in Lexington: ComAir Flight 5191 Took Off From Wrong Runway

08/27/2006: Airline disasters in Kentucky are rare; today’s crash of ComAir Flight 5191 is in fact the worst such crash in Kentucky history. The total death toll is 49. The first officer–James M. Polehinke–is the only survivor. He is in critical condition at University of Kentucky Hospital.

According to WHAS in Lexington, it has been confirmed that the plane–which crashed just after takeoff–took off from the wrong runway, and that there was one contoller in the tower at the time.

This could either be controller error–in which the contoller provided the wrong instruction to the pilot–or the pilot failing to follow directions according to the takeoff checklist. A serious issue will be whether takeoff clearance from that runway was requested and/or granted.

Unforunately, the runway used for takeoff was 3,500 feet; however, according to the most immediate information I am finding, the CRJ-200 needs 5,800 feet to take off when fully-loaded. (WHAS is saying 5,009 feet, but I’m seeing 5,800. However, that difference is moot, as–either way–there is a catastrophic variance between takeoff distance required and runway length.)

The plane was at almost full capacity–it seats 50. At this point, there has been no indication that there was an engine failure which, on takeoff, would have been a double whammy.

UPDATE: It is also being reported–by WHAS in Lexington–that there was a small amount of rain at the time, and that there was some downdraft. Whether or not the strength of this would have been sufficient to bring down an aircraft has not been confirmed. However, combined with a short runway, the flight may not have had sufficient airspeed to withstand even a mild wind disturbance.

UPDATE 2: According to CNN, the flight was cleared for takeoff on runway 4-22, which was 7,000 feet long. Instead, the crew used runway 8-26. The latter was for general aviation–small craft–and only 3,500 feet in length.

Possible factors: crew fatigue, failure by crew to relay takeoff instructions in the checklist, failure by controller to alert aircraft that wrong runway is being used (this is possible if that lone controller was directing incoming aircraft).

UPDATE 3: That report of rain may have been inaccurate, as–according to the National Weather Service in Louisville–winds were only about 10 mph and the rain had cleared out at 6 am, the approximate time of takeoff. There were no visibility restrictions and no clouds.

This is looking more and more like pilot error due to failure to follow portions of the checklist. Takeoff instructions to the crew from the controller would have been detailed down to the turn directions, and the pilot is responsible for confirming to ensure that the proper runway is being used before taking off. In this case, the ultimate buck stops with the pilot-in-command.
The crew had ample flight experience, including with the CJ-200. Then again, a pilot with lots of experience flying early in the morning could easily be tempted not to use the checklist. Failure to do that is a very common cause for accidents on takeoff and landing. One of the worst of which was Northwest Airlines Flight 255 from Detroit (August 16, 1987). This killed 148 passengers, 6 crew, and two motorists on I-94. A 4-year-old girl, shielded from the flames by her mother, survived.

Correction: Cecilia Cichan–the lone survivor of Northwest Airlines Flight 255–was not shielded by her mother (as was reported at the time…I remember those like it happened yesterday!). What makes her survival all the more miraculous is that her mom was found several feet from where Cecilia was found.

UPDATE 4: WHAS in Lexington is reporting that Doppler images from this morning are in fact indicating that there was a small storm that quickly dissipated, which happened to be right at the spot where the crash occurred. (This would contradict the initial reports out of the National Weather Service in Louisville.) Winds were estimated at about 20 mph at the time. The role–if any–that may have played in the accident is only ascertainable from the flight data recorders. We await the official word from the NTSB.

A 20 mph wind may not seem like much, but if an airplane is struggling to gain altitude after taking off from a runway that was too short, even such an otherwise innocuous shift of wind could have dropped that plane right into the trees.

So far, the confirmed dead include:

(1) Horse trainer Jeff Williams

(2) Jonathan Walton Hooker–a former UK baseball player–and his wife, Scarlet Catherine Parsley. They were just married last night.

(3) Patrick Smith, a member of the International Board of Habitat for Humanity. He was heading for Gulfport, Mississippi for the anniversary of the Hurricane Katrina disaster.

(4) Jeffrey Clay, the Captain. A Vineland, New Jersey native, he was not far from my former stomping grounds (the ‘burbs of Philadelphia).

(5) Kelly Heyer, the flight attendant.

The really good news: Hats off to the first responders–who included a Lexington Police Officer–who were on the scene almost immediately, who pulled First Officer James M. Polehinke from the wreckage. Hopefully, he will make it.

The really bad news (and there will be lots of it as the cases unfold): A 16-year-old girl was killed on the flight. Her mom–who was supposed to be on that flight as well–was bumped at the last minute. I’ll be praying for that poor mom who now must bury her 16-year-old child.

UPDATE 5: The NTSB is saying that this was a CRJ-100, not a CRJ-200. The former is the older version of the same aircraft. However, the takeoff distance seems to be the same. Converting the given distance–1,768 meters–to feet gives about 5,800 feet. Ergo, the smaller runway still would have been equally problematic.

UPDATE 6: The NTSB (and WHAS Lexington) are reporting that the aircraft crashed through the perimeter fence. This is significant in that it would have been at least as contributory to the accident as any speculative wind gust. This is because the fence could have caused structural damage to the wings, control surfaces, and also the engine that would have hindered the flight from gaining altitude.

One respondent asked if the crew could have tried to abort the takeoff once they realized they were on the wrong runway. While that is possible, that would have been an extremely high-risk maneuver, as given the shorter runway–the “decision speed”; i.e., the speed past which aborting the takeoff is not possible–would be lower than for the longer runway.

I would suggest that–more than likely–they attempted to take off because aborting would have been more certain disaster given the shorter reaction time required for aborting. Aborting a takeoff on such a short runway–with a jet aircraft–would have been very difficult even if the crew had recognized their error early. Short runway=RAZOR THIN margin of error.

UPDATE 7: Another factor that has not been mentioned as a contributor is the possibility that the plane could have been overloaded. The CRJ-100 seats 50, and there were 47 passengers and three crew (50 people). Depending on the total weight of the luggage–and the total weight of the passengers–it is possible that the plane may have been overloaded.

Coupled with the takeoff on a short runway, if this were the case it would have been an aggravating factor.

In aerospace terms, overloading causes the center of gravity to fall behind the “aerodynamic center”, which makes the aircraft “statically unstable”.

In layman’s terms, it means you will be absolutely unable to fly the plane safely.

Ultimately, the loading of the aircraft is the responsibility of the pilot-in-command, but even he or she is at the mercy of the available instrumentation and/or their estimation skills. Depending on the instrumentation, that part is not always an exact science.

It’s easy to make all those hypothetical weight and balance calculations at a desk; it’s another matter when you have real loading situations at 6 in the morning on a Sunday.

UPDATE 8: In a briefing currently in progress, the NTSB indicated that when the aircraft began acceleration, “it continued accelerating until it crashed”. That would indicate that there was no attempt to abort the takeoff. (I expected that because–given that the runway was too short–such an evasive move would have carried even more risk than attempting to complete the takeoff once any error might have been recognized. Once the takeoff roll had commenced, the crew probably had three seconds or less to decide to abort the takeoff.)

UPDATE 9: According to radio station WVLK in Lexington, the crew had arrived in Lexington at about 12:30 am on Sunday. If this account is confirmed as fact, then crew fatigue is going to be a huge contributory factor.

Given that Flight 5191 crashed at 6:07 am, the crew likely had less than 4 hours of sleep. Such fatigue has been known to cause flight crews to get lax with their checklists, and this case is almost clearly a failure to execute the checklist. (The runway number includes the heading; the controller had provided the correct heading; and the crew failed to confirm before takeoff that the plane was pointed at the correct heading– that is a checklist issue.)

American Airlines Flight 1420 (Little Rock, AR, 01 June 1999) is a prominent example of this, as crew fatigue was listed as a contributory factor as to why the pilots failed to ensure that the spoilers were extended after the aircraft had touched down. (That was a failure to execute the checklist.)

Leave a Reply

Your email address will not be published.

Connect with Facebook

This site uses Akismet to reduce spam. Learn how your comment data is processed.