NTSB ADVISORY
National Transportation Safety Board
Washington, DC 20594
December 15, 2005
NTSB UPDATE ON SOUTHWEST AIRLINES RUNWAY OVERRUN AT MIDWAY
AIRPORT
The National Transportation Safety Board today released
the following update on its investigation into the accident
involving Southwest Airlines flight 1248, a Boeing 737-700 on
December 8, 2005, at Midway Airport in Chicago, Illinois. The
airplane overran runway 31C during the landing rollout.
The accident occurred about 7:14 pm central standard
time. The airplane departed the end of the runway, rolled
through a blast fence, a perimeter fence, and onto a
roadway. The airplane came to a stop after impacting two
automobiles. One automobile occupant was fatally injured
and another seriously injured. The flight was conducted
under 14 CFR Part 121 and had departed from the
Baltimore/Washington International Thurgood Marshall
Airport, Maryland.
The on-scene portion of the investigation has been
completed. Additional fact-finding, including tests and
research, will be conducted at various component
manufacturers. The Safety Board staff continues to examine
the information provided by the flight data recorder and the
cockpit voice recorder.
Operations/Human Performance
The two pilots in the cockpit were interviewed on
Saturday. Each interview took approximately three hours.
The pilots stated that everything was normal through
the point of touchdown. Approaching the airport, weather
was of concern to them, and they listened to the ATIS (the
recorded weather update) four times during the latter
portion of the flight. They stated that they agreed with
the dispatcher's assessment of the conditions for landing on
runway 31C and backed up that assessment by inputting the
numbers into the on-board laptop computer tool.
The computer confirmed that the landing would be
within the operational parameters of the airplane and
Southwest's procedures, they said. Autobrakes were set on
MAX, and they activated after a "firm" touchdown. The
flying pilot (Captain) stated that he could not get the
reverse thrust levers out of the stowed position. The first
officer, after several seconds, noticed that the thrust
reversers were not deployed and activated the reversers
without a problem. At some point, the Captain noticed that
the airplane was not decelerating normally and applied
maximum braking manually. The first officer also became
aware of the poor braking effectiveness, moved his seat
farther forward, and also applied maximum braking. They
stated that they continued to apply maximum pressure to the
brakes as the airplane went straight off the end of the
runway and came to a stop.
Interviews were conducted with a number of other
Southwest Airlines flight crews, including the crew of the
last Southwest flight to land at Midway and a subsequent
crew that diverted to St. Louis.
Airplane Performance
Preliminary calculations show that the airplane
touched down with about 4,500 feet of remaining runway and
was on the runway for about 29 seconds. Preliminary
calculations also show that, for the runway conditions and
use of brakes and thrust reverser that occurred, the
stopping distance without hitting obstructions would have
been about 5,300 feet (the actual stopping distance was
about 5,000 feet). In addition, had the airplane landing
into the wind, rather than with a tail wind, the stopping
distance for a landing would have been about 1,000 feet
less.
Documentation of aircraft performance from the scene
has been completed to the maximum extent possible. It was
not possible to observe tire marks from much of the landing
rollout due to the fact that the aircraft landed on a snow-
covered runway and snow fell on the runway immediately
following the accident.
FDR data show that autobrakes were active and provided
high brake pressure upon touchdown. Autobrakes and manual
braking continued to provide high brake pressure throughout
the landing roll.
FDR data show that thrust reversers were activated
about 18 seconds after touchdown or about 14 seconds before
contact with the blast fence. Testing and examination of
the thrust reverser systems will continue.
Investigators have obtained the laptop computer tool
used by the accident flight crew. It will be examined and
calculations of landing performance will be compared to
flight manual data.
Eleven security-type video cameras were identified on
the airport that may show imagery of the airplane rollout or
the surface of the runway and taxiway at the time of the
accident. The videos will be reviewed.
Meteorology
National Weather Service forecasters and other
personnel were interviewed. An enhanced snow band was in
the area producing localized heavy snow due to lake
enhancement. This apparently is a somewhat unusual weather
phenomenon, as the band swath was only 20 to 30 miles wide
with snow accumulations of 10 inches right over Midway
Airport.
Midway Airport weather observation equipment and
records were examined and all equipment was working normally
during the evening of the accident.
Southwest Airlines dispatchers who were associated
with the accident flight were interviewed. Prior to the
takeoff from Baltimore, when weather conditions deteriorated
and the runway switched to runway 31C, the dispatcher
determined that runway 31C was approved for landing for
flight 1248. Runway conditions, braking action, wind speed
and direction, airplane weight and mechanical condition of
the aircraft are typical factors considered in making such
decisions. The flight was contacted twice on the way to
Midway and the appropriateness of using the runway for
landing was reaffirmed during both contacts.
Official weather observations:
Approximately 20 minutes prior to the accident, the
winds were from 100 degrees at 11 knots, visibility was «
mile in moderate snow and freezing fog, the ceiling was
broken at 400 feet, and overcast at 1400 feet, temperature -
3C, dew point -5C, altimeter setting 30.06 in. Hg. Remarks -
runway 31C rvr (runway visual range) 4500 feet, snow
increment - 1 inch of new show in the last hour, 10 inches
on the ground.
Approximately 23 minutes after the accident, a special
observation revealed winds out of 160 degrees at 5 knots,
visibility ¬ mile in heavy snow, freezing fog, sky obscured
with a vertical visibility of 200 feet, temperature -4C,
dewpoint -5C. Remarks - runway 31C, rvr 3000 feet.
Toxicology
Blood and urine samples were obtained from both
pilots. The disposition of the blood samples is being
reviewed.
Structures
The aircraft has been removed from the accident site
and was transferred to a hangar at Midway Airport. The
maintenance log revealed no writeups or deferred items for
the accident flight or several previous flights.
Professional surveyors completed a survey of the
accident scene and the geography leading up to the site to
include the locations of parts shed by the aircraft after it
left the paved runway surface and the blast fence destroyed
during the accident sequence.
Powerplants
Both engines were visually examined at the accident
site. Although the first stage compressor blades of both
engines showed foreign object damage, they were all intact
and present. Wood from the blast fence and other debris was
present in both engines. A visual examination of the
turbine sections revealed no missing blades.
The 60-day engine history revealed no deferrals or
writeups. Each engine has two thrust reverser sleeves. FDR
data indicated that all four sleeves were deployed until
after the airplane left the paved runway overrun surface.
Hydraulic system B (that runs the thrust reversers) revealed
no leaks.
Systems
The Systems Group documented the switches, circuit
breakers and controls in the cockpit. The leading edge
slat, flap, and trailing edge flap extension measurements
were taken and revealed symmetrical extension of all
devices. The measurements will be compared to Boeing
documentation to determine exact extension.
Chicago Fire Department personnel were interviewed to
determine if any switch positions or other items were
altered during the rescue effort. The Fire Department Chief
stated that the only things his people did were to
disconnect the battery and turn off the crew oxygen source.
The brakes were found in good condition with adequate
wear remaining. The main landing gear tires had acceptable
tread depth and no indication of flat spots.
Air Traffic Control
The local controller, two tower controllers, and the
tower supervisor were interviewed. All controllers stated
that they saw the aircraft lights during the landing roll,
but did not see the actual touchdown.
The investigation has revealed that runway 31C was
used as the landing runway because it contained lower
landing minimums for aircraft using the ILS approach. If
runway 13C was used, the runway most aligned with the wind,
pilots would have been unable to land because of
insufficient landing minimums.
Survival Factors/Airports
All flight attendants were interviewed. They all said
that they noted a smooth landing but that the deceleration
feeling thereafter seem less than usual. They noted that
the emergency lighting came on after the airplane came to
rest, and one flight attendant opened the L1 door to begin
the evacuation. The emergency slide deployed automatically,
but its angle in relation to the ground was less than ideal.
This caused passengers to begin to pile up around the
bottom of the slide. Rescue personnel assisted people away
from the slide. The first officer deplaned after about 5
passengers and also assisted in getting people away from the
airplane.
Further factual updates will be issued when
appropriate.
NTSB Media Contact: Keith Holloway, 202-314-6100
Thanks.
I was trying to put myself in the mind of the flight crew that night. No CAT III, so screw that idea.