Crash During Approach to Landing Empire Airlines Flight 8284

CrashDuringApproachtoLanding

EmpireAirlinesFlight8284

AvionsdeTransportRégional

AerospatialeAleniaATR42‐320,N902FX

Lubbock,Texas

January27,2009



Accident Report

NTSB/AAR-11/02

PB2011-910402

National

Transportation

Safety Board

NTSB/AAR-11/02

PB2011-910402

Notation 8093A

Adopted April 26, 2011

Aircraft Accident Report

Crash During Approach to Landing

Empire Airlines Flight 8284

Avions de Transport Régional

Aerospatiale Alenia ATR 42-320, N902FX

Lubbock, Texas

January 27, 2009

National

Transportation

Safety Board

490 L’Enfant Plaza, S.W.

Washington, D.C. 20594

National Transportation Safety Board. 2011. Crash During Approach to Landing, Empire Airlines

Flight 8284, Avions de Transport Régional Aerospatiale Alenia ATR 42-320, N902FX, Lubbock,

Texas, January 27, 2009. Aircraft Accident Report NTSB/AAR-11/02. Washington, DC.

Abstract: This accident report discusses the January 27, 2009, accident involving Empire Airlines

flight 8284, an Avions de Transport Régional Aerospatiale Alenia ATR 42-320, N902FX, which crashed

short of the runway at Lubbock Preston Smith International Airport, Lubbock, Texas. The captain

sustained serious injuries, and the first officer sustained minor injuries. The airplane was substantially

damaged. The airplane was registered to FedEx Corporation and operated by Empire Airlines, Inc., as a

14 Code of Federal Regulations Part 121 supplemental cargo flight. Instrument meteorological conditions

prevailed, and an instrument flight rules flight plan was filed. The safety issues discussed in this report

include the flight crew’s actions in response to the flap anomaly, the continuation of the unstabilized

approach, the dispatch of the flight into freezing drizzle conditions, the efficiency of the emergency

response, and simulator-based training for pilots who fly in icing conditions. Nine safety

recommendations are addressed to the Federal Aviation Administration.

The National Transportation Safety Board is an independent Federal agency dedicated to promoting aviation,

railroad, highway, marine, pipeline, and hazardous materials safety. Established in 1967, the agency is mandated by

Congress through the Independent Safety Board Act of 1974 to investigate transportation accidents, determine the

probable causes of the accidents, issue safety recommendations, study transportation safety issues, and evaluate the

safety effectiveness of government agencies involved in transportation. The Safety Board makes public its actions

and decisions through accident reports, safety studies, special investigation reports, safety recommendations, and

statistical reviews.

Recent publications are available in their entirety on the Internet at <http://www.ntsb.gov>. Other information about

available publications also may be obtained from the website or by contacting:

National Transportation Safety Board

Records Management Division, CIO-40

490 L’Enfant Plaza, SW

Washington, DC 20594

(800) 877-6799 or (202) 314-6551

Safety Board publications may be purchased, by individual copy or by subscription, from the National Technical

Information Service. To purchase this publication, order report number PB2011-910402 from:

National Technical Information Service

5285 Port Royal Road

Springfield, Virginia 22161

(800) 553-6847 or (703) 605-6000

The Independent Safety Board Act, as codified at 49 U.S.C. Section 1154(b), precludes the admission into evidence

or use of Board reports related to an incident or accident in a civil action for damages resulting from a matter

mentioned in the report.

NTSB Aircraft Accident Report

Contents

Figures ........................................................................................................................................... iv

Abbreviations and Acronyms .......................................................................................................v

Executive Summary ..................................................................................................................... ix

1. Factual Information ...................................................................................................................1

1.1 History of Flight .........................................................................................................................1

1.2 Injuries to Persons ......................................................................................................................5

1.3 Damage to Airplane ...................................................................................................................5

1.4 Other Damage ............................................................................................................................5

1.5 Personnel Information ................................................................................................................6

1.5.1 The Captain ......................................................................................................................6

1.5.2 The First Officer ..............................................................................................................6

1.5.3 Flight Crew’s Rest Opportunities ....................................................................................7

1.5.3.1 The Captain ....................................................................................................... 7

1.5.3.2 The First Officer ............................................................................................... 8

1.6 Airplane Information .................................................................................................................8

1.6.1 General .............................................................................................................................8

1.6.2 Flap System .....................................................................................................................9

1.6.3 Deicing, Anti-Icing, and Anti-Icing Advisory Systems ................................................11

1.6.4 Flight Instruments ..........................................................................................................11

1.6.4.1 Airspeed Indicator ........................................................................................... 11

1.6.4.2 Electronic Flight Information System ............................................................. 13

1.6.5 Automatic Flight Control System ..................................................................................13

1.6.6 Stall Protection System ..................................................................................................13

1.7 Meteorological Information .....................................................................................................14

1.8 Aids to Navigation ...................................................................................................................14

1.9 Communications ......................................................................................................................15

1.10 Airport Information ................................................................................................................15

1.10.1 Emergency Plan ...........................................................................................................15

1.10.2 Snow and Ice Control Plan ..........................................................................................16

1.11 Flight Recorders .....................................................................................................................16

1.11.1 Cockpit Voice Recorder ..............................................................................................16

1.11.2 Flight Data Recorder ....................................................................................................17

1.12 Wreckage and Impact Information ........................................................................................17

1.13 Medical and Pathological Information ...................................................................................18

1.14 Fire .........................................................................................................................................18

1.15 Survival Aspects ....................................................................................................................18

1.15.1 Flight Crew Egress ......................................................................................................18

1.15.2 Emergency Response ...................................................................................................18

1.16 Tests and Research .................................................................................................................21

1.16.1 Airplane Performance Study and Simulator Evaluations ............................................21

NTSB Aircraft Accident Report

1.16.1.1 Effects of Flap Asymmetry ........................................................................... 21

1.16.1.2 Effects of Airframe Ice Accretion ................................................................. 22

1.16.2 Terrain Awareness and Warning System ....................................................................23

1.16.3 Flap System Component Examinations .......................................................................24

1.16.3.1 Hydraulic Fluid Testing ................................................................................ 25

1.16.3.2 Maintenance History ..................................................................................... 25

1.16.3.3 Flap Asymmetry Event Involving Another Airplane .................................... 25

1.16.4 Deicing and Anti-Icing System Examinations ............................................................26

1.16.5 Autopilot Disconnect Alert ..........................................................................................26

1.17 Organizational and Management Information .......................................................................26

1.17.1 Pilot Standard Operating Procedures and Training .....................................................27

1.17.1.1 Flap Anomalies ............................................................................................. 27

1.17.1.2 Stabilized Approach and Go Around ............................................................ 31

1.17.1.3 Terrain Awareness and Warning System ...................................................... 32

1.17.1.4 Airspeed Bugs ............................................................................................... 32

1.17.1.5 Aileron Mistrim with Autopilot Engaged ..................................................... 33

1.17.1.6 Deicing and Anti-Icing Systems ................................................................... 33

1.17.1.7 Crew Resource Management ........................................................................ 34

1.17.1.8 Weather Information and Training ............................................................... 35

1.17.1.9 Stall and Near-Stall Recoveries .................................................................... 36

1.17.2 Aircraft Deicing Program (Ground) ............................................................................36

1.17.3 Flight Dispatch .............................................................................................................36

1.17.4 Hazardous Materials ....................................................................................................38

1.17.5 Federal Aviation Administration Oversight ................................................................38

1.17.6 Previous Accidents ......................................................................................................39

1.18 Additional Information ..........................................................................................................40

1.18.1 Postaccident Safety Action ..........................................................................................40

1.18.1.1 Federal Aviation Administration .................................................................. 40

1.18.1.2 Empire Airlines ............................................................................................. 41

1.18.1.3 FedEx Feeder Operations .............................................................................. 42

1.18.2 Previously Issued Safety Recommendations ...............................................................43

1.18.2.1 Flight Crew Monitoring and Workload Management Skills for Single and

Multiple Abnormal Situations..................................................................................... 43



1.18.2.2 Go-Around Callout ....................................................................................... 45

1.18.2.3 Low-Airspeed Alerting Systems ................................................................... 45

1.18.2.4 First Officer Assertiveness ............................................................................ 46

1.18.3 Airplane Equipment Options .......................................................................................47

1.18.3.1 Aircraft Performance Monitoring System .................................................... 47

1.18.3.2 Flap Malfunction or Asymmetry Light ......................................................... 48

1.18.3.3 Ice Evidence Probe ....................................................................................... 49

2. Analysis .....................................................................................................................................50

2.1 General .....................................................................................................................................50

2.2 Accident Sequence ...................................................................................................................52

2.3 Human Performance ................................................................................................................54

2.3.1 Captain’s Nonstandard Responses .................................................................................55

2.3.2 Flight Crew’s Failure to Monitor Airspeed ...................................................................57

NTSB Aircraft Accident Report

2.3.2.1 Airspeed References ....................................................................................... 58

2.3.2.2 Flight Crew Monitoring Skills and Techniques .............................................. 60

2.3.3 Crew Resource Management .........................................................................................61

2.3.4 Go-Around Callout ........................................................................................................63

2.3.5 Training for Single and Multiple Abnormal Situations Occurring at Low Altitude .....64

2.3.6 Fatigue Considerations ..................................................................................................65

2.3.6.1 Role of Fatigue in Flight Crew Performance .................................................. 67

2.4 Operations Issues .....................................................................................................................69

2.4.1 Dispatch into Freezing Drizzle Conditions ....................................................................69

2.4.1.1 Flight Crewmember and Dispatcher Training Related to Icing Conditions ... 70

2.4.1.2 Proposed Rulemaking for Airplane Certification Requirements for Flight in

Icing Conditions .......................................................................................................... 72

2.4.2 Training Program and Manual Discrepancies ...............................................................72

2.5 Survival Aspects ......................................................................................................................73

2.5.1 Lack of Occupant and Hazardous Materials Information Available

to First Responders .................................................................................................................73

2.5.2 Inoperative Emergency Response and Mutual Aid Gate ...............................................75

2.6 Airplane Equipment Options ...................................................................................................75

2.6.1 Aircraft Performance Monitoring ..................................................................................75

2.6.2 Flap Malfunction or Asymmetry Light ..........................................................................77

2.6.3 Ice Evidence Probe ........................................................................................................77

2.7 Simulator Training for Icing Encounters .................................................................................78

3. Conclusions ...............................................................................................................................80

3.1 Findings....................................................................................................................................80

3.2 Probable Cause.........................................................................................................................83

4. Recommendations ....................................................................................................................84

Board Member Statements .........................................................................................................86

5. Appendixes ...............................................................................................................................93

Appendix A: Investigation and Hearing ........................................................................................93

Appendix B: Cockpit Voice Recorder Transcript ..........................................................................94

iii

NTSB Aircraft Accident Report

Figures

Figure 1. Aerial view of approach end of runway 17R and location of wreckage. ....................... 5

Figure 2. Flap position indicator. Inset shows the likely needle pointer position during the

flap asymmetry.............................................................................................................................. 10



Figure 3. Exterior flap position fairing. ....................................................................................... 10

Figure 4. Airspeed indicator showing needle pointer; selector knob; and internal, yellow,

white, and red bugs. ...................................................................................................................... 12

Figure 5. View of airport showing relative locations of airport access gates and accident site. . 20

Figure 6. Right side of fuselage showing extensive fire damage to right wing and

flap components. ........................................................................................................................... 24

Figure 7. Quick reference handbook page 2.21 with “FLAPS JAM/UNCOUPLED/ASYM” and

“REDUCED FLAPS LANDING” procedures. ............................................................................ 28

Figure 8. General information on page 0.02 of quick reference handbook. ................................ 29

NTSB Aircraft Accident Report

Abbreviations and Acronyms

AAS anti-icing advisory system

AC advisory circular

AD airworthiness directive

ADP aircraft deicing program

ADU advisory display unit

AEP airport emergency plan

AFFF aqueous film forming foam

AFM airplane flight manual

AFW Fort Worth Alliance Airport

agl above ground level

AIM Aeronautical Information Manual

AIRMET airmen’s meteorological information

AOA angle of attack

APM aircraft performance monitoring

ARAC Aviation Rulemaking Advisory Committee

ARFF aircraft rescue and firefighting

ATC air traffic control

ATCT air traffic control tower

ATOS air transportation oversight system

ATR Avions de Transport Régional

ATR 42 Aerospatiale Alenia ATR 42-320

BEA Bureau d’Enquêtes et d’Analyses

CFR Code of Federal Regulations

NTSB Aircraft Accident Report

CG center of gravity

CRM crew resource management

CVR cockpit voice recorder

DGAC Direction Générale de l’Aviation Civile

EADI electronic attitude direction indicator

EASA European Aviation Safety Agency

EFS engineering flight simulator

ELP El Paso International Airport

EMS emergency medical service

FAA Federal Aviation Administration

FAF final approach fix

FCOM flight crew operating manual

FDR flight data recorder

FedEx FedEx Corporation

FL flight level

fpm feet per minute

FR Federal Register

FSDO flight standards district office

FSIB flight standards information bulletin

FTM flight training manual

GOM general operations manual

HAZMAT hazardous materials

Hg mercury

IEP ice evidence probe

NTSB Aircraft Accident Report

IFR instrument flight rules

ILS instrument landing system

IMC instrument meteorological conditions

InFO information for operators

KIAS knots indicated airspeed

kts knots

LBB Lubbock Preston Smith International Airport

lbs pounds

LOM locator outer marker

LWD left-wing-down

MAF Midland International Airport

msl mean sea level

NASA National Aeronautics and Space Administration

NPRM notice of proposed rulemaking

NTSB National Transportation Safety Board

NWS National Weather Service

PF pilot flying

PIC pilot-in-command

PM pilot monitoring

PMI principal maintenance inspector

POI principal operations inspector

QRH quick reference handbook

RWD right-wing-down

SAFO safety alert for operators

SIC second-in-command

vii

NTSB Aircraft Accident Report

SIGMET significant meteorological information

SLD supercooled large droplet

SN serial number

SOP standard operating procedure

STC supplemental type certificate

SVFR special visual flight rules

TAWS terrain awareness and warning system

TSB Transportation Safety Board of Canada

VFR visual flight rules

minimum approach airspeed for 30° flaps plus 5 kts (not corrected for wind) or

1.1 V

mca

, whichever is greater

mca

minimum control airspeed with 5° of bank and the failure of the critical engine

with takeoff flaps and landing gear retracted

mHB

30 minimum approach airspeed and target touchdown airspeed with 30° flaps

mLB

minimum approach airspeed for 0° flaps

VMC visual meteorological conditions

viii

NTSB Aircraft Accident Report

Executive Summary

On January 27, 2009, about 0437 central standard time, an Avions de Transport Régional

Aerospatiale Alenia ATR 42-320, N902FX, operating as Empire Airlines flight 8284, was on an

instrument approach when it crashed short of the runway at Lubbock Preston Smith International

Airport, Lubbock, Texas. The captain sustained serious injuries, and the first officer sustained

minor injuries. The airplane was substantially damaged. The airplane was registered to FedEx

Corporation and operated by Empire Airlines, Inc., as a 14 Code of Federal Regulations Part 121

supplemental cargo flight. The flight departed from Fort Worth Alliance Airport, Fort Worth,

Texas, about 0313. Instrument meteorological conditions prevailed, and an instrument flight

rules flight plan was filed.

The National Transportation Safety Board determines that the probable cause of this

accident was the flight crew’s failure to monitor and maintain a minimum safe airspeed while

executing an instrument approach in icing conditions, which resulted in an aerodynamic stall at

low altitude. Contributing to the accident were 1) the flight crew’s failure to follow published

standard operating procedures in response to a flap anomaly, 2) the captain’s decision to

continue with the unstabilized approach, 3) the flight crew’s poor crew resource management,

and 4) fatigue due to the time of day in which the accident occurred and a cumulative sleep debt,

which likely impaired the captain’s performance.

The safety issues discussed in this report include the flight crew’s actions in response to

the flap anomaly, the continuation of the unstabilized approach, the dispatch of the flight into

freezing drizzle conditions, the efficiency of the emergency response, and simulator-based

training for pilots who fly in icing conditions. Nine safety recommendations are addressed to the

Federal Aviation Administration.

NTSB Aircraft Accident Report

1. Factual Information

1.1 History of Flight

On January 27, 2009, about 0437 central standard time,

an Avions de Transport

Régional (ATR) Aerospatiale Alenia ATR 42-320 (ATR 42), N902FX, operating as Empire

Airlines flight 8284, was on an instrument approach when it crashed short of the runway at

Lubbock Preston Smith International Airport (LBB), Lubbock, Texas. The captain sustained

serious injuries, and the first officer sustained minor injuries. The airplane was substantially

damaged. The airplane was registered to FedEx Corporation (FedEx) and operated by Empire

Airlines, Inc., as a 14 Code of Federal Regulations (CFR) Part 121 supplemental cargo flight.

The flight departed from Fort Worth Alliance Airport (AFW), Fort Worth, Texas, about 0313.

Instrument meteorological conditions (IMC) prevailed,

and an instrument flight rules (IFR)

flight plan was filed.

During the flight’s initial descent to LBB, the first officer was the pilot flying (PF) with

the autopilot system engaged. As the airplane descended from 14,000 to 8,000 feet above mean

sea level (msl), the captain, who was the pilot monitoring (PM), performed the descent and

approach checklists. Review of the cockpit voice recorder (CVR) transcript revealed that the

captain confirmed that the airplane’s anti-icing and deicing protection was set to level 3.

The

captain performed the approach briefing,

which included what he believed to be the “icing

speed” information (minimum maneuvering and operating airspeeds for approach in icing

conditions).

According to the CVR transcript, at 0419:41, the captain briefed the first officer about the

missed approach procedure for the instrument landing system (ILS) approach to runway 17R,

stating that “in the event of a miss it’ll be uh climb to thirty seven and a right turn to five

All times in this report are central standard time (unless otherwise noted) and based on a 24-hour clock.

According to the Federal Aviation Administration (FAA) Pilot/Controller Glossary, IMC is defined as

“meteorological conditions expressed in terms of visibility, distance from cloud, and ceiling [that are] less than the

[minimums] specified for visual meteorological conditions.” At the time of the accident airplane’s approach, LBB’s

ceiling and visibility were below the minimums specified in 14 CFR 91.155 for flight under visual flight rules.

Information obtained from the FAA’s website <http://www.faa.gov/air_traffic/publications/atpubs/PCG/> (accessed

March 21, 2011).

The airplane flight manual contained guidelines for using the anti-icing and deicing systems. For more

information, see section 1.17.1.6.

During postaccident interviews, the flight crewmembers reported that the captain performed the approach

briefing because they anticipated that he would be flying the approach. They believed that weather conditions were

at the minimums for the approach (ceiling 200 feet above ground level, visibility 1/2 mile), and they were unsure if

the first officer had accumulated the 100 hours of flight time in the ATR 42 required by Empire Airlines for her to

fly the approach to minimums. Subsequently, they learned that the weather conditions were above minimums and

decided that the first officer would fly the approach.

For more information about the captain’s approach airspeed briefing and the flight crew’s procedures for

obtaining and using such information, see section 1.17.1.4.

According to the CVR transcript, the air traffic controller informed the flight crew that runway 8/26 was

closed and that the localizer back course approach, which the first officer recognized as the only approach for

runway 35L, was unavailable. The flight crew discussed that the winds were from the north but that runway 17R

was the only available option.

NTSB Aircraft Accident Report

hundred feet

[7]

via the Lubbock one fourteen radial to HYDRO intersection and hold annnd

[8]

that looks like a parallel entry.” The first officer acknowledged this statement, and the captain

then stated, “climb to thirty seven and a left turn to five out to one fourteen.”

At 0422:32, the LBB approach controller cleared the flight to descend to 6,000 feet msl

and advised that he had not received any icing reports and that the special weather observation

about 0408 indicated winds from 350° at 10 knots (kts), visibility 2 miles in light freezing drizzle

and mist, ceiling overcast at 500 feet above ground level (agl), temperature -8° C, dew point

-9° C, and altimeter setting 30.12 inches of mercury (Hg). At 0424:57, the captain contacted

company flight operations on the radio to advise that the flight was about 15 minutes out.

At 0433:13, the controller cleared the flight for the ILS approach to runway 17R, and at

0433:52, he instructed the flight to contact the LBB tower. The captain established contact with

the tower, and, at 0434:06, the tower controller cleared the flight to land and reported that the

wind was from 010° at 8 kts.

During postaccident interviews, both the captain and the first officer stated that the

airplane accumulated ice during the descent. At 0434:24, the first officer called for a flap setting

of 15°, the deployment of the landing gear, and the landing check. During a postaccident

interview, the first officer recalled that the airplane subsequently accelerated in the descent,

which she stated was unusual for the airplane in a flaps- and gear-down configuration. A

postaccident review of information from the airplane’s flight data recorder (FDR) showed that,

after the first officer called for flaps, the airplane’s airspeed remained about 158 kts and briefly

increased to about 160 kts, and the first officer decreased engine power to about 3 percent torque.

At 0435:03, about the time that the first officer reduced engine power, she stated, “what the heck

is going on?” At 0435:04, the captain replied, “you know what? we have no flaps.”

The FDR showed that a flap asymmetry had occurred in which the right flaps did not

extend and the left flaps extended partially (8° to 10°).

The autopilot countered the flap

asymmetry by applying aileron deflection that resulted in a 20° left control-wheel movement.

The FDR indicated that, at the time the captain stated that they had “no flaps,” the airplane was

at an altitude of about 1,400 feet agl just outside the locator outer marker (LOM), which serves

as the final approach fix (FAF).

According to the CVR transcript, after the captain commented about the flap problem,

neither crewmember discussed a procedure or checklist to address it. During a postaccident

interview, the first officer stated that she continued to fly the approach with the airplane coupled

to the autopilot while the captain got out a flashlight and tried to troubleshoot the flap problem.

The published procedure specified that the initial climb to 3,700 feet msl is followed by a climbing “left” turn

to “5,000 [feet msl]” via the radial to the intersection.

The use of excess vowels, letters, or drawn-out syllables in any word in the CVR transcript is intended to

provide a phonetic representation of the word as spoken.

According to the ATR 42 Airplane Flight Manual, the maximum allowable tailwind component for landing is

15 kts. The wind recorded about 0430 by the LBB automated surface observing system was from 360° at 14 kts

gusting to 19 kts.

All references in this section to airplane configuration, airspeed, altitude agl, and engine power settings are

derived from the airplane performance study. For more information about the study, see section 1.16.1.

NTSB Aircraft Accident Report

The captain stated in a postaccident interview that he repositioned the flap handle several times

and used the flashlight to check the circuit breakers behind the first officer’s seat. The captain

stated that, after finding that no circuit breakers were out, he moved the flap handle back to the

“up” (or 0°) position because he did not want the flaps to travel inadvertently during the

approach.

The FDR indicated that, between 0434:58 and 0435:25, as the airplane descended with

the autopilot engaged, its airspeed decreased from about 160 to 125 kts. At 0435:30 (about

26 seconds after the captain commented about the flap anomaly), the CVR captured a sound

consistent with the aural stall warning (“cricket”) and the tactile stick shaker that lasted

1.1 second. The captain stated to the first officer, “yeah don’t do that.” At 0435:32, a 0.3-second

sound consistent with the aural stall warning sounded. At 0435:36, the captain stated, “just keep

flying the airplane. okay.” The FDR indicated that the stick-shaker activation disconnected the

autopilot when the airplane was about 900 feet agl at 125 kts indicated airspeed (KIAS).

The

FDR also indicated that the first officer increased engine power to about 70 percent torque and

began manually flying the approach. Following the application of power, the airspeed began

increasing. At 0435:40, the first officer asked, “should I go around,” and the captain replied,

“no,” and then stated, “keep descending.”

At 0435:44, the first officer stated, “we’re getting close here”; the CVR transcript

indicated that the first officer’s voice sounded as if she were straining. During a postaccident

interview, the captain stated that he looked over and was surprised to see that the first officer was

manually flying the approach and that he had not heard the aural alert to indicate that the

autopilot had been disconnected (no alert was captured by the CVR). The captain asked the first

officer if she wanted him to finish the approach; the first officer replied, “yes please.”

At 0435:50, the captain took control as the PF when the airplane was about 700 feet agl at

143 KIAS. During a postaccident interview, the captain stated that the airplane was high and to

the right of the approach course and that he corrected to put it back on course. According to the

FDR, at 0435:56, the captain reduced the engine power to about 10 percent torque. At 0436:00,

the CVR captured a 0.9-second sound consistent with the aural stall warning and the stick shaker

concurrent with the terrain awareness and warning system (TAWS) warning, “pull up. pull up.”

According to the FDR, at this time, the airplane was about 500 feet agl at 156 KIAS. At 0436:04,

the first officer stated, “there’s the runway.”

Between 0436:00 and 0436:17, as the airplane continued to descend, its airspeed

decreased from about 156 to 129 kts. At 0436:17, the captain called for maximum propeller

speed. At 0436:19, when the airplane was about 200 feet agl at 124 KIAS, the CVR captured the

sounds of engine power increasing concurrent with the aural stall warning and the stick shaker.

At 0436:21, the airplane was about 150 feet agl at 123 KIAS and began to roll right-wing-down

(RWD).

Between 0436:21 and 0436:23, the airplane rolled right to a bank angle of 35° RWD. The

airplane then rolled left to a bank angle of 50° left-wing-down (LWD) bank, and then right

The published minimum airspeed for the airplane flying with the flaps retracted in icing conditions was

143 KIAS.

NTSB Aircraft Accident Report

reaching a 10° RWD bank before it impacted the ground. At 0436:27, the CVR captured the

sound of impact. During a postaccident interview, the captain reported that he had no lateral

control of the airplane and that the controls were almost “snatched” out of his hands.

An airport operations agent in a vehicle on taxiway M, which parallels runway 17R/35L,

was about 1,800 feet south of the approach end of runway 17R when he saw the airplane

approach. He estimated that the airplane was about 400 to 500 feet agl when it emerged from the

fog layer and stated that it appeared to be high and to the right of the runway centerline. He

stated that the airplane appeared to be “pancaking” in its descent and that it rolled slightly to the

right, then rolled hard to the left in a near-90° bank, then rolled back to the right, impacted the

ground, and slid for a distance. He further stated that, as the airplane slid past his position, a fire

erupted on the airplane’s right wing and “engulfed” the airplane. He stated that he immediately

notified his supervisor of the accident and that he could hear the beeping tones of the crash

phone over his radio as the airplane continued to slide. He stated that he soon saw the lights of

the aircraft rescue and firefighting (ARFF) units emerging from their station.

The airplane came to rest next to the right side of the runway (about 200 feet west of the

runway centerline), north of taxiway S. Figure 1 is an aerial view of the approach end of

runway 17R and the location of the wreckage. The ARFF units, which were on the accident

scene within about 4 minutes, were the first responders to arrive. The ARFF response was joined

by Lubbock Fire Department units, one of which could not enter at the designated airport access

gate because the gate would not open; the unit was diverted to a different gate to enter the

airport.

For more information about the emergency response, see section 1.15.2.

NTSB Aircraft Accident Report

Figure 1. Aerial view of approach end of runway 17R and location of wreckage.

1.2 Injuries to Persons

Table. Injury Chart

Injuries Flight Crew Cabin Crew Passengers Other Total

Fatal

0 0 0 0 0

Serious

1 0 0 0 1

Minor

1 0 0 0 1

None

0 0 0 0 0

Total

2 0 0 0 2

1.3 Damage to Airplane

The airplane was substantially damaged by the impact forces and postcrash fire.

1.4 Other Damage

The airport approach lighting system and the runway were damaged during the accident

sequence.

NTSB Aircraft Accident Report

1.5 Personnel Information

1.5.1 The Captain

The captain, age 52, held an airline transport pilot certificate with ratings for

single-engine and multiengine land airplanes. His most recent Federal Aviation Administration

(FAA) first-class medical certificate was issued on September 19, 2008, with a limitation that he

must “possess glasses for near/intermediate vision.” According to Empire Airlines’ records, the

captain had accumulated 13,935 total flight hours, including 12,742 hours as pilot-in-command

(PIC). He had a total time in the ATR 42 of 2,052 hours, 1,896 hours of which were as PIC. In

the 90 days, 30 days, and 24 hours before the accident, the captain accumulated about 53, 13, and

5 hours, respectively.

The captain was hired by Empire Airlines on May 9, 1988. Employee training records

showed that the captain had previously flown and served as a check airman on the company’s

Cessna 208 (single-engine, turbopropeller-powered) airplanes, in which he had accumulated

about 6,600 total flight hours. He had also previously flown the company’s Fokker F27

(multiengine, turbopropeller-powered) airplanes, in which he had accumulated about 2,500 total

flight hours, about 1,900 hours of which were as PIC. Employee training records showed that the

captain completed his most recent recurrent proficiency check ride and line check on

September 22, 2008. His most recent recurrent ground training, which focused on ATR 42

systems, was completed on March 29, 2008. A review of FAA and company records revealed

that the captain had no previous FAA enforcement actions, incidents, or accidents.

The captain was experienced with in-flight icing conditions because he had worked as a

pilot in the Pacific Northwest and Alaska for 30 years. He stated that he had been dispatched into

freezing drizzle before and that, while flying in such conditions, he maintained a heightened

awareness of the flying environment.

The first officer described the captain as someone whom she would feel comfortable

speaking up to, if necessary. She stated that the captain sought her input and asked for her

opinion and that, at the beginning of their trip pairing, the captain told her that she should tell

him if she saw him do anything wrong. Other Empire Airlines’ pilots who had flown with the

captain stated that his greatest strength as a pilot was his experience level. One pilot stated that

the captain occasionally cut corners, seemed rushed, was less thorough on briefings than other

captains, and became easily agitated when flying and driving.

1.5.2 The First Officer

The first officer, age 26, held a commercial pilot certificate with ratings for single-engine

and multiengine land and instrument airplanes. Her most recent FAA first-class medical

certificate was issued on December 4, 2008, with no limitations. According to Empire Airlines’

records, the first officer had accumulated 2,109 total flight hours, about 1,890 hours of which

were as PIC. She had accumulated 130 hours in the ATR 42 as second-in-command (SIC). In the

90 days, 30 days, and 24 hours before the accident, the first officer accumulated about 88, 29,

and 5 hours, respectively.

NTSB Aircraft Accident Report

The first officer was hired by Empire Airlines on July 25, 2008. Employee training

records showed that the first officer completed her most recent proficiency check ride (an initial

check) on September 10, 2008. Her most recent ground training, which was initial training for

the ATR 42, was completed on August 29, 2008. A review of FAA and company records

revealed that the first officer had no previous FAA enforcement actions, incidents, or accidents.

The first officer had limited experience flying in icing conditions before working for

Empire Airlines, and the ATR 42 was the first airplane in which she had flown that was equipped

with deicing and anti-icing systems.

Other Empire Airlines pilots who had flown with the first officer described her flying

skills as “at par” or “average” compared to other first officers with the same experience level.

One pilot stated that the first officer was methodical about using checklists. Another pilot stated

that the first officer’s greatest strength as a pilot was that she was “nonconfrontational.” One

captain described the first officer’s crew resource management (CRM) skills as “good,” and

another captain stated that the first officer did not seem to have a problem speaking up, if

needed. Regarding areas in which the first officer could improve, one pilot stated that “she could

employ the skills she already knew without asking so many questions,” and another pilot stated,

“more hands-on flying of the airplane.”

1.5.3 Flight Crew’s Rest Opportunities

1.5.3.1 The Captain

The captain reported that, on January 24, 2009, he awoke about 0800. From about 1015

to 1535, he traveled as a passenger on commercial flights to reposition from his home in

Portland, Oregon, to Midland International Airport (MAF), Midland, Texas. The captain went to

bed about 2200 in his hotel room in Midland, Texas. On January 25, 2009, the captain awoke

about 0800, performed various activities throughout the day (including exercising, reading,

shopping, and going to dinner at a coworker’s home with the first officer), and went to bed about

2200.

The captain stated that, on January 26, 2009, he awoke about 0400 to prepare for the

upcoming night flight. He performed various activities and then slept from about 1100 to 1630.

He stated that he met the first officer in the hotel lobby about 1820 and that they drove to MAF

to begin duty about 1845. They departed MAF about 1945, arriving at El Paso International

Airport (ELP), El Paso, Texas, about 2115. They subsequently departed ELP for AFW about

2230, arriving about 0030 on January 27, 2009. The flight crew then departed AFW for LBB (the

accident flight) about 0313.

During postaccident interviews, the captain indicated that he needed 6 to 8 hours of sleep

per night to feel rested. He stated that he considered himself to be a night person and that he felt

rested before the accident flight. He described his workload on the day of the accident as normal

overall but stated that, during the approach, the workload was high. The captain stated that

takeoffs and landings were high-workload situations and that the icing conditions and flap

anomaly elevated his workload.

NTSB Aircraft Accident Report

1.5.3.2 The First Officer

The first officer (who was based in Salt Lake City, Utah)

had first arrived in Midland,

Texas, on January 18, 2009, to fly a trip sequence with another captain and had completed that

trip sequence on January 23, 2009. During those trips, the first officer was on duty during night

hours beginning about 0045. After the first officer completed that trip sequence, Empire Airlines

paid the expenses for the first officer to spend the weekend in Midland, Texas, rather than have

her commute home on a short turnaround. The first officer indicated that she was off duty from

the afternoon of January 24, 2009, until the evening of the accident flight. The first officer

indicated that, because her previous trip sequence and the accident trip sequence required her to

sleep during the day and be awake at night, she maintained that same sleep schedule during her

off-duty time (between the two trip sequences) in Midland, Texas.

On January 24, 2009, the first officer went to bed about 0600 in her hotel room and

awoke about 1430. She then performed various activities throughout the day and evening

(including exercising and going to the store) and went to bed about 0600 on January 25, 2009.

She awoke about 1400, later met the captain to go to dinner at a coworker’s home, returned to

the hotel about 2200, and went to bed sometime in the morning on January 26, 2009. The first

officer awoke about 1500 and reported for her trip duty with the captain about 1845.

During postaccident interviews, the first officer indicated that she needed 7 hours of sleep

to feel rested. She stated that she considered herself to be a night person and that she felt rested

before the accident flight. She described her workload on the day of the accident as normal but

stated that the workload became high when she and the captain realized that the flaps did not

function properly. The first officer stated that although the workload during the approach was

high, she felt that it was something that they could handle.

1.6 Airplane Information

1.6.1 General

The accident airplane (model ATR 42-320, serial number [SN] 175) received its

airworthiness certificate in 1990. In 2005, the airplane was configured for freight operations in

accordance with supplemental type certificate (STC) ST01189W1.

The airplane was powered

by two Pratt & Whitney Canada PW 121 engines, each equipped with a Hamilton Standard 14SF

propeller. According to the type certificate data sheet, the ATR 42 is certificated for flight in

known icing conditions in accordance with 14 CFR Part 25, Appendix C, which details the types

of icing exposures for which the aircraft is tested and certificated. The airplane’s most recent

inspection (a 2-month inspection) was completed on January 9, 2009. At the time of the accident,

the airplane had accumulated 28,768 hours of operating time and 32,379 cycles.

Salt Lake City, Utah, was a “floater” base for Empire Airlines. Flight crews based there would commute to

various cities based on the company’s needs and at the company’s expense.

STC ST01189W1 is held by M7 Aerospace LP.

NTSB Aircraft Accident Report

The airplane had no deferred maintenance items, and the maintenance records showed

compliance with all applicable airworthiness directives (ADs) for the airframe, propeller

subassemblies, and the right engine.

The AD tracking form for the left engine was not located.

During the investigation, an Empire Airlines records clerk located work orders and records of

other required directives to demonstrate AD compliance for the left engine.

The airplane’s maximum gross takeoff and landing weights were 37,258 and

36,155 pounds (lbs), respectively. A review of the FedEx Feeder Aircraft Load Control Sheet

and the Empire Airlines Cargo Load Manifest for the flight revealed that the accident flight’s

calculated takeoff and estimated landing weights were 34,487 and 32,717 lbs, respectively, and

that its calculated center of gravity (CG) was within limits.

1.6.2 Flap System

The ATR 42 is equipped with four trailing-edge flaps (two per wing) that are electrically

controlled and hydraulically operated (each flap has a hydraulic actuator). The flight crew can

command the flaps into four positions using the cockpit flap control lever: 0° for cruise flight,

15° for takeoff and approach, 30° for landing, and 45° for use during an emergency.

During

normal operations, all flaps move together when commanded.

An interconnection torque shaft is connected to the left and right inboard flaps. Both a

flap position transmitter, which provides flap position information to the flap indicator in the

cockpit, and a flap asymmetry detector are attached to the interconnection torque shaft. The flap

position transmitter is mounted near the center of the interconnection torque shaft and senses the

average position of the flaps based on movement of the torque shaft. The flap asymmetry

detector is designed to prevent any asymmetry between the right and left flaps from exceeding 8°

to 10° (the precise value depends upon the flap positions when the asymmetry occurs). If a flap

asymmetry reaches the determined value, a microswitch integrated into the detector cuts off the

electrical power supply to the flap control switch, and the flap extension or retraction solenoid is

no longer energized. As a result, the flaps will stop in the positions reached at the time of the

power interruption. Once this occurs, the flaps will not move in response to movement of the flap

control lever in the cockpit until maintenance personnel reset the system on the ground.

If a flap asymmetry occurs due to a restriction (jam) and the restriction is then removed,

hydraulic pressure in the system will move the flaps to a symmetrical position. The resulting flap

position will be the average of the right and left flap positions when the asymmetry occurred.

Even if such hydraulic balancing of the flaps occurs, post-flight maintenance action would still

be required to restore electrical power to the flap control switch.

The airframe AD compliance records were located throughout the airplane’s records, the propeller

subassembly ADs were tracked through Empire Airlines’ computerized maintenance tracking system, and the right

engine ADs were listed on an AD tracking form located with the right engine logbooks.

The four flap positions are standard increments. In the ATR 42, for some selected settings, the actual position

of the flaps differs by a few degrees from the indicator increment (the 30° flap position increment represents 27° of

actual flap deflection).

For more information about the flight crew’s procedures for responding to a flap asymmetry, see

section 1.17.1.1.

NTSB Aircraft Accident Report

A flap position indicator in the cockpit has an index marked with the four flap positions

(shown as 0°, 15°, 30°, and 45°), and a needle pointer indicates the flap position on the index. If

a flap asymmetry occurs, the flap indicator needle in the cockpit will point to the index value that

corresponds with the average position of the flaps. Figure 2 shows the flap position indicator,

and the inset depicts the likely needle pointer position during the flap asymmetry.

Figure 2. Flap position indicator. Inset shows the likely needle pointer position during the flap

asymmetry.

In addition to the cockpit flap position indicator, exterior flap position fairings mounted

on the wings (one per wing) are visible to each flight crewmember through their respective side

windows. These fairings are also marked with lines and numbers for the four flap positions,

providing flight crews with a visual indication of the actual position of each flap. Figure 3 shows

the exterior flap position fairing. The wing flap position fairings are illuminated by the wing

lights during nighttime use. If a flap asymmetry occurs, the actual positions of the right and left

flaps will be indicated on the respective flap position fairings.

Figure 3. Exterior flap position fairing.

NTSB Aircraft Accident Report

1.6.3 Deicing, Anti-Icing, and Anti-Icing Advisory Systems

The ATR 42’s ice protection system permits the airplane to operate in some atmospheric

icing conditions (applicable restrictions are noted in the limitations section of the airplane flight

manual [AFM]).

The leading edges of the ATR 42’s wings, horizontal stabilizers, and vertical

stabilizer are equipped with pneumatic deice boots that are inflated by engine compressor bleed

air.

When the flight crew selects airframe pneumatic boot use, pockets in the boot system

inflate in sequence, perpendicular to the airfoil, to deice the leading edges.

The propeller blades, windshields, probes (pitot, static, air temperature, and

angle-of-attack [AOA]), and flight control horns

are electrically heated. The horn anti-icing

feature prevents ice deposits from forming between the wing structure and the moving parts of

the control surfaces. A pneumatic system provides deicing protection for the engine intakes and

is designed to prevent a reduction or total loss of engine performance in icing conditions.

According to the ATR 42 flight crew operating manual (FCOM), the ATR 42 has an

anti-icing advisory system (AAS) designed to alert the flight crew to apply the correct

procedures when flying in icing conditions.

The AAS includes an ice detector (mounted under

the left wing) and three cockpit light annunciators; the amber ICING light and green ICING

AOA light are on the central panel, and the blue DEICING light is on the memo panel. The AAS

alerts the flight crew as soon as (and as long as) the ice detector detects ice accretion. Once the

probe detects at least 0.5 millimeter of ice, the amber ICING light illuminates. The AAS

self-tests continuously, and any failure illuminates a FAULT light with a single chime.

1.6.4 Flight Instruments

1.6.4.1 Airspeed Indicator

The ATR 42 is equipped with three analog airspeed indicators (one on each pilot’s

instrument panel and one on the central panel). Each airspeed indicator has a moving needle

pointer that indicates the airplane’s KIAS on a fixed scale that shows 60 to 400 kts. Each

airspeed indicator has four moveable indices, known as “bugs” (one internal and three external),

on the outer scale that the pilot uses to manually set certain airspeed references for takeoff or

landing, based on airplane performance information for the flight conditions.

The internal bug, which is underneath the glass face of the airspeed indicator, is set by the

pilot using a selector knob. According to the ATR 42 pilot handbook, for approach and landing

in icing conditions, the internal bug should be set to V

mHB

30 icing, which is the minimum

For more information about these restrictions, see section 1.17.1.6.

The vertical stabilizer deice boots are optional on ATR 42s.

The airplane’s primary flight controls (ailerons, elevators, and rudder) are mechanically operated and have

counterweights, or horns, that serve to balance the controls.

For information about flight crew procedures for responding to each AAS annunciator, see section 1.17.1.6.

For information about obtaining applicable airspeed references and setting the bugs, see section 1.17.1.4.

NTSB Aircraft Accident Report

approach airspeed and target touchdown airspeed with 30° flaps in icing conditions.

The

airspeed indicated by the internal bug (as selected by the pilot using the selector knob) controls

the reference on the electronic attitude direction indicator (EADI) fast/slow scale.

The other three indices, which are located on the outside of the glass face of the airspeed

indicator, are color coded for use in referencing certain airspeeds during takeoff or landing and

are manually positioned by the pilot. Unlike the internal bug’s function, the pilot’s positioning of

these three indices (yellow, white, and red) does not affect any other systems or displays. The

ATR 42 pilot handbook states that, for an approach, the yellow bug should be set to reference

, which is the approach airspeed for 30° flaps plus 5 kts (not corrected for wind effect).

Similarly, the white bug should be set to reference V

mLB

0 normal, which is the minimum

airspeed for an approach with 0° flaps in nonicing conditions, and the red bug should be set to

reference V

mLB

0 icing, which is the minimum airspeed for an approach with 0° flaps in icing

conditions. Figure 4 is an airspeed indicator showing the needle pointer; selector knob; and

internal, yellow, white, and red bugs.

Figure 4. Airspeed indicator showing needle pointer; selector knob; and internal, yellow, white,

and red bugs.

The ATR 42 pilot handbook notes that this speed is to be corrected for wind and that the wind factor is 1/3 of

the headwind velocity or the gust in full, with a maximum wind factor of 15 kts.

For more information, see section 1.6.4.2.

The ATR 42 pilot handbook states that for V

, pilots should use the higher of V

mHB

30 + 5 kts (not corrected

for wind) or 1.1 V

mca

(1.1 times the minimum control airspeed with 5° of bank and the failure of the critical engine

with takeoff flap and landing gear retracted).

NTSB Aircraft Accident Report

1.6.4.2 Electronic Flight Information System

The airplane’s electronic flight information system includes displays for each pilot on the

cockpit instrument panel. Each pilot’s display includes an EADI to indicate the airplane’s pitch

and roll relative to the horizon. The EADI depicts the horizon as a straight, horizontal line

between the blue upper half of the display (showing the relative location of the sky), and the

brown lower half (showing the relative location of the ground). The pitch scale is marked in 5°

intervals, and the roll scale indicates select bank angles between 0° and 60°.

The left side of each EADI also displays a “fast/slow” scale that indicates the difference

between the airplane’s actual airspeed and the target airspeed as selected by the pilot using the

internal bug on the airspeed indicator. An index on the fast/slow scale moves up (fast) or down

(slow) according to any deviation of the actual airspeed from the selected airspeed.

1.6.5 Automatic Flight Control System

The airplane was equipped with an automatic flight control system that provided

autopilot, yaw damper, flight director, and altitude alert functions. The autopilot system is

designed to automatically disconnect when certain system conditions occur, including achieving

the stall warning threshold. If an autopilot disconnect occurs (either automatically or manually),

indications provided to the flight crew include an aural alert (“cavalry charge” tone),

a message

(“AP DISENGAGED”) on the advisory display unit (ADU), and a visual annunciator (“AP

MSG”) on the primary flight display. With the autopilot system engaged, if an anomaly in the

airplane’s roll characteristics occurs (such as those that a flap asymmetry would induce), the

ADU will show “RETRIM ROLL R [or L] WING DN” and “AILERON MISTRIM” messages.

1.6.6 Stall Protection System

The airplane’s stall protection system includes two AOA probes (one on each side of the

forward fuselage) that detect the airplane’s AOA, an aural stall warning (“cricket”), a tactile stick

shaker, and a stick pusher. According to ATR, when the airplane is configured for flight in

nonicing conditions with flaps 0°, if the AOA probes detect that the airplane has reached a

critical AOA of 11.6°, the aural stall warning sounds, and the stick shaker activates. If the probes

detect that the AOA continues to increase, the stick pusher activates between 11.5° and 13.8°,

depending on the AOA’s rate of change.

The stall protection system is designed to provide an earlier stall warning threshold

during flight in icing conditions. According to ATR, when the airplane is configured with the

control horns’ anti-icing selected on with flaps 0°, the critical AOA for activation of the aural

stall warning and the stick shaker is reduced to 7°, and the stick-pusher activation remains at an

angle between 11.5° and 13.8°.

The aural warning can be inhibited by other aural warnings with higher priority.

NTSB Aircraft Accident Report

1.7 Meteorological Information

Freezing precipitation in the form of light freezing rain and ice pellets began at LBB

about 2300 on January 26, 2009. The precipitation later changed to light freezing drizzle, which

continued until after the accident.

The LBB weather observation about 0353 reported wind from 010° at 14 knots, visibility

3 miles in light freezing drizzle and mist, ceiling overcast at 500 feet agl, temperature -8° C, dew

point -9° C, and an altimeter setting 30.13 inches of Hg. The LBB weather observation about

0453 reported wind from 020° true at 13 kts gusting to 18 kts, visibility 2 miles in light freezing

drizzle and mist, ceiling overcast at 500 feet agl, temperature -8° C, dew point -9° C, and an

altimeter setting 30.13 inches of Hg. The remarks section of the weather observation indicated

that the ceiling was variable from 400 to 900 feet agl and that the hourly liquid equivalent

precipitation was less than 0.01 inch (a trace of precipitation).

The National Weather Service (NWS) terminal aerodrome forecast for LBB expected

IFR conditions to prevail during the period including the accident flight’s arrival with wind from

030° at 12 knots, visibility of 1 1/2 miles in light freezing drizzle and mist, and ceiling overcast

at 100 feet agl through about 0800, with a temporary period of visibility 1/2 mile in light freezing

drizzle and freezing fog. The NWS also had issued an airmen’s meteorological information

(AIRMET) for moderate icing conditions

from the freezing level through 22,000 feet msl over

Texas that was current at the time of the accident. A review of the NWS current icing potential

and supercooled large droplet (SLD) icing diagnostic charts indicated a greater-than-80-percent

probability of icing conditions over the LBB area from the surface through 6,000 feet msl and a

high likelihood of encountering SLD conditions. Temperatures were above freezing between

6,000 and 11,000 feet msl as a result of a temperature inversion.

1.8 Aids to Navigation

No problems with any navigational aids were reported.

As defined by the American Meteorological Society Glossary and the National Weather Service Federal

Meteorological Handbook, freezing drizzle is drizzle that falls in liquid form but freezes upon impact with surfaces

to form a coating of glaze; freezing drizzle has a diameter between 0.05 to 0.5 millimeters (0.002 to 0.02 inches) at

temperatures less than 0° C.

According to the FAA’s Aeronautical Information Manual (AIM), when reporting icing to air traffic control,

pilots should describe the icing intensity as moderate when the rate of accumulation is such that even short

encounters become potentially hazardous and use of deicing/anti-icing equipment or flight diversion is necessary.

Information obtained from the FAA’s website <http://www.faa.gov/air_traffic/publications/ATpubs/AIM/>

(accessed March 21, 2011). FAA Advisory Circular 91-74A, “Pilot Guide: Flight In Icing Conditions” published in

December 2007, states that, in icing of moderate intensity, “the rate of ice accumulation requires frequent cycling of

manual deicing systems to minimize ice accretions on the airframe. A representative accretion rate for reference

purposes is 1 to 3 inches (2.5 to 7.5 cm) per hour on the outer wing. The pilot should consider exiting the condition

as soon as possible.”

NTSB Aircraft Accident Report

1.9 Communications

No problems with any communications equipment during the flight were reported.

During a postaccident interview, the first officer stated that, after the accident, the captain

attempted to contact the air traffic control tower (ATCT) using the airplane’s radios but that the

radios were inoperative. The CVR transcript indicates that the CVR captured the sound of impact

at 0436:27, a subsequent sound similar to occupants moving around in the cockpit, the captain’s

instructions about exiting the airplane, and the sound of the door opening at 0437:25. No further

sounds were captured, and the recording ended at 0441:35.

1.10 Airport Information

LBB is located about 4 miles north of the city of Lubbock, Texas, at an elevation of

3,282 feet msl. LBB is a Class I public airport certificated under 14 CFR Part 139.

Runway

17R/35L, which is the longest of the airport’s three runways,

has a concrete surface and is

11,500 feet long and 150 feet wide.

LBB maintained ARFF capabilities at Index C

and had four vehicles, three of which

(Rescue 1, 2, and 4) were equipped with a roof turret, bumper turret, and hand line; each vehicle

carried 1,500 gallons of water, at least 190 gallons of aqueous film forming foam (AFFF), and

450 lbs of dry chemical. The remaining vehicle (Rescue 5), which was equipped with a hand

line, carried 450 lbs of dry chemical and 100 gallons of a mixture of water and AFFF. The ARFF

station was centrally located south of taxiway J near the midpoint of runway 8/26. ARFF

services were available 24 hours a day, every day, and the station was staffed by a minimum of

four firefighters. Five firefighters, all of whom participated in the response, were on duty at the

time of the accident.

1.10.1 Emergency Plan

LBB had an FAA-approved airport emergency plan (AEP) that provided guidance

intended to minimize the possibility and extent of personal injury and property damage during an

emergency. The FAA approved the most recent revisions to the AEP on July 9, 2008. The AEP

stated that, during an emergency response, “primary entry points to the airfield by responding

units will be at Gate 6 and Gate 48 or at a location prescribed by the command post.” The AEP

stated that “the Battalion Chief will set up the initial staging area/command post near Gate 6,

Gate 48, or another designated area.”

Title 14 CFR 139.5 states that a Class I airport is certificated to serve scheduled operations of large air carrier

aircraft and can also serve unscheduled passenger operations of large air carrier aircraft and/or scheduled operations

of small air carrier aircraft.

Runway 8/26, which has a concrete surface, is 8,001 feet long and 150 feet wide, and runway 17L/35R,

which has an asphalt surface, is 2,891 feet long and 75 feet wide.

LBB was an Index C airport based on five or more average daily departures of aircraft with a length of at

least 126 feet but less than 159 feet, as defined in 14 CFR 139.315. For an airport to meet Index C capabilities,

14 CFR 139.317 requires two or three ARFF vehicles that contain at least the minimum specified amount of various

extinguishing agents.

For more information about the emergency response, see section 1.15.2.

NTSB Aircraft Accident Report

If an accident occurs, the ATCT notifies the ARFF station through the direct line radio to

the station. According to the letter of agreement between the LBB ATCT and the City of

Lubbock, upon the alert activation, the ATCT controller should provide, “as is available,” the

“type of aircraft, nature of the emergency, runway to be used, and any other information if time

permits.”

1.10.2 Snow and Ice Control Plan

As required by 14 CFR 139.313, LBB maintained an FAA-approved snow and ice

control plan. According to the airport’s snow and ice operations document, the operations

supervisor, operations agent on duty, or maintenance supervisor is responsible for monitoring the

runway and taxiway environment to determine when snow and ice removal is needed. Sensors

embedded at various points on the runways, taxiways, and the terminal ramp detect surface and

subpavement temperatures and surface moisture, and information from the sensors is monitored

from the airfield maintenance building by airport maintenance personnel.

The snow and ice operations document indicated that, during critical temperatures, if

moisture or precipitation is occurring or is expected to occur, granulated urea will be applied to

the first priority airport operational surfaces, such as the active runway, its parallel taxiway,

taxiways connecting the active runway to the terminal ramp, and the established ARFF road

from the fire station to taxiway J. The assistant airport manager stated that the road from the fire

station to taxiway J had received a urea application before the accident flight’s arrival. No

written record exists to indicate when this application occurred.

The snow and ice operations plan did not address monitoring or ensuring the operability

of the airport emergency response and mutual aid gates. FAA Advisory Circular (AC)

150/5200-30C, “Airport Winter Safety and Operations,” which provides guidance to assist

airport operators in developing a snow and ice control plan, also did not address this topic. An

informal survey of other 14 CFR Part 139 airports where operations are affected by winter

weather revealed that none of the airports’ snow and ice control plans addressed gate operability.

1.11 Flight Recorders

1.11.1 Cockpit Voice Recorder

The accident airplane was equipped with a Fairchild Model A100A CVR, SN 59653,

capable of recording four channels of analog audio on a continuous loop tape. An examination of

the CVR at the National Transportation Safety Board’s (NTSB) vehicle recorders laboratory

found no evidence of structural or heat damage, and audio information was extracted without

difficulty.

The extracted 30-minute, 43-second recording consisted of three channels of useable

audio information that captured events from before the flight’s descent to several minutes after

the accident. Excellent quality audio

was obtained from the channels for the captain’s and the

In an excellent quality recording, virtually all of the crew conversations can be accurately and easily

understood, and only one or two words are not intelligible; any loss is usually attributed to simultaneous cockpit

NTSB Aircraft Accident Report

first officer’s audio panels, and good quality audio was obtained from the channel for the cockpit

area microphone. The fourth channel (the use of which was not required by Federal regulations)

contained no audio information.

1.11.2 Flight Data Recorder

The accident airplane was equipped with a Fairchild Model FA2100 FDR, SN 289112.

The FDR was sent to the NTSB’s laboratory for readout and evaluation; it was received in good

condition, and the data were extracted normally from the recorder. The FDR recorded about

56 hours of data, and the accident flight is reflected in the final 2 hours 18 minutes of the

recording. The FDR system was configured to record more than 70 parameters of airplane flight

information in a digital format using solid-state flash memory as the recording medium. For this

investigation, 58 parameters that were considered relevant were verified and examined. The

relevant parameters included flap position, flap asymmetry, left and right aileron surface

positions, KIAS, glideslope and localizer deviations, radio altitude (altitude as sensed by the

radio altimeter), AOA, roll, pitch, pitch trim, right control column position, control wheel

position, airframe deicing, left and right elevator positions, and left and right engine propeller

speed.

1.12 Wreckage and Impact Information

Examination of the accident site showed that the airplane initially impacted flat, grassy

terrain about 300 feet north of the runway 17R blast pad.

Ground scars and debris from the

airplane extended about 2,500 feet south from that location to where the main wreckage came to

rest next to the right side of the runway (about 200 feet west of the runway centerline), north of

taxiway S. Fragments of the nose landing gear, right main landing gear, and a separated outboard

section of the right aileron were identified in the initial debris field near the beginning of the

runway.

Examination of the wreckage found that the airplane was resting on its nose section and

leaning to the right; the left main landing gear remained attached. Sections of the upper fuselage,

right wing, and right inboard and outboard flaps were destroyed by fire. The left engine remained

secured to the wing, and the left propeller sustained minimal damage. The right engine and the

right propeller sustained impact and fire damage.

conversations and/or radio transmissions that obscure each other. In a good quality recording, most of the crew

conversations can be accurately and easily understood, but several words or phrases are not intelligible; any loss can

be attributed to minor technical deficiencies, momentary dropouts in the recording system, or a large number of

simultaneous cockpit conversations and/or radio transmissions that obscure each other.

For information regarding the relevant parameters that were used in performing the airplane performance

study, see section 1.16.1.

A blast pad, which is unusable for landing, takeoff, or taxiing, is designed to eliminate the erosive effect of

the high wind forces produced by airplanes at the beginning of their takeoff roll. The blast pad for runway 17R is a

concrete extension north of the threshold that can also serve as a stopway in the event of a rejected takeoff on

runway 35L.

NTSB Aircraft Accident Report

1.13 Medical and Pathological Information

On the morning of January 27, 2009, the captain and the first officer participated in

breath analysis testing for alcohol and provided urine specimens for company drug testing.

The

tests yielded no evidence of alcohol or drug use by either crewmember.

1.14 Fire

A fuel-fed fire erupted during the accident sequence and destroyed sections of the

airplane and some cargo before being suppressed by firefighting personnel. According to the

airplane’s dangerous goods manifest, 25 lbs of dry ice, one package of seatbelt pretensioners,

and two containers of materials classified as “ORM-D [other regulated materials-domestic]

consumer commodity” were destroyed in the fire. Other hazardous material (HAZMAT) on

board, which included five containers of liquid chloride, one container of liquid sodium chlorite,

and one container of solid sodium iodine (all radioactive materials), were undamaged.

1.15 Survival Aspects

1.15.1 Flight Crew Egress

According to the CVR transcript, after the airplane came to a stop, the captain told the

first officer to “go out the hatch” but then stated, “there’s a fire on the right hand side. go out the

left.” During a postaccident interview, the first officer stated that she first opened the cockpit

escape hatch and saw a “raging fire” on the airplane’s right side. She stated that she then went to

the forward cargo door, unlocked it, and pushed it open with the help of the captain. The first

officer stated that they “jumped out and ran to FedEx.” The captain stated that, once they were

outside the airplane and a safe distance from it, he called Empire Airlines’ dispatch to report the

accident. Both the captain and the first officer left the accident scene before the first responders

arrived.

1.15.2 Emergency Response

The ATCT notified the ARFF station of the accident through the direct line radio.

According to LBB ARFF personnel, a tone sounded in the fire station, followed by “muffled

voices.” Shortly thereafter, a second tone sounded, and an Alert 3 (which indicates an accident

on or near the airport) was announced. According to the dispatch logs from the Lubbock Fire

Department, the alarm sounded at 0438:12, and three ARFF units (Rescue 2, 4, and 5) were

dispatched at 0438:46 and arrived on scene at 0442:31.

The ARFF captain, who was driving Rescue 5, reported that the three ARFF units

proceeded along taxiway J toward taxiway M. He stated that ice was on the runways and taxiway

and that he proceeded with caution because of the potential for slippery conditions. The ARFF

equipment operator driving Rescue 2, which also carried an ARFF lieutenant and another

Empire Airlines’ drug testing program checked the samples provided for evidence of marijuana, cocaine,

amphetamines, opiates, and phencyclidine use.

NTSB Aircraft Accident Report

firefighter, estimated that they proceeded no faster than 45 mph. While en route to the accident

site, the ARFF equipment operator driving Rescue 2 contacted the ATCT controller to ask what

type of airplane had crashed and how many passengers were on board; the ATCT controller

replied that it was an ATR carrying an unknown number of people.

The ARFF lieutenant in Rescue 2 stated that the darkness and fog made it difficult at

times to distinguish between smoke and fog. The ARFF captain driving Rescue 5 stated that,

after turning onto taxiway M, he saw both smoke and the glow from the fire. The ARFF captain

stated that heavy smoke and flames engulfed the airplane from the cockpit to the wing section

and extended just aft of the wing, where most of the flames were present. The ARFF equipment

operator driving Rescue 2 stated that they immediately began applying foam and water to the

wreckage and, within about 45 seconds, suppressed the fire to the extent that they could see the

FedEx logo on the airplane’s tail. He indicated that, because of the logo, he knew that it was a

cargo airplane and that he would expect two to three occupants. He stated that he then heard

Rescue 5’s call to start looking for survivors.

The ARFF captain stated that the fire was “blackened down” quickly. The ARFF

lieutenant from Rescue 2 accessed the cabin area from a burn hole in the fuselage to look for

occupants, and another firefighter from Rescue 2 used a ladder to access the cockpit roof hatch to

look inside. After observing that no one was inside the airplane, ARFF personnel from Rescue 2

and Rescue 5 immediately began searching the area for the airplane’s occupants by performing

visual and thermal imaging searches.

According to standard protocol for an Alert 3 accident, Lubbock Fire Department units

also responded and were dispatched at 0441:10. The response included the battalion chief and

engine from Station 2 (which is about 3.2 miles from LBB and staffed with personnel who are all

ARFF-certified firefighters with airport access badges); an engine from Station 5 (about 7 miles

from LBB); and an engine, truck, and HAZMAT team from Station 4 (about 7 miles from LBB).

As directed, one engine from Station 2 responded to Gate 48, which is an electronically

controlled gate operated by a drive chain. The engine team was unable to enter the airport

property at Gate 48 because the gate had ice in its operating mechanism and would not open. The

Station 2 engine then drove to the ARFF station to proceed from that point. All other responding

Lubbock Fire Department units were informed that the wreckage was closest to the FedEx

hangar and were rerouted to respond to that site. A FedEx employee who opened the gate near

the FedEx hangar for the Station 2 team, Station 5 engine, and emergency medical service

(EMS) personnel informed them that the pilots were at the hangar; the EMS units proceeded to

the hangar. Figure 5 is a view of the airport showing the relative locations of airport access gates

and the accident site.

NTSB Aircraft Accident Report

Figure 5. View of airport showing relative locations of airport access gates and accident site.

The first Lubbock Fire Department unit arrived on scene at 0457:27. At that time, the

ARFF units had contained the fire, with the exception of a few deep-seated cargo fires inside the

airplane. The Lubbock Station 2 battalion chief assumed incident command and informed the

ARFF captain that the flight crew was at the FedEx hangar. The ARFF captain stopped the

search for survivors, and ARFF personnel resumed their fire suppression activities. The Station 2

battalion chief and ARFF personnel indicated that the ATCT controller later called and stated

that a pilot was walking around on the ramp looking for assistance, and a second search for

survivors was initiated. ARFF personnel determined that both pilots were being transported to a

local hospital, and the emphasis on fire suppression resumed.

The Lubbock battalion captain (who assumed command of operations from the Lubbock

Station 2 battalion chief) stated that, about 5 to 10 minutes after he arrived on scene, the

Lubbock Police Department delivered a HAZMAT manifest that FedEx employees had

provided. The fire department had not requested the information. The Lubbock Station 2

battalion chief stated that the HAZMAT team responded to the FedEx hangar to obtain the

material safety data sheets for the materials on the HAZMAT manifest. The ARFF captain stated

that he did not see the HAZMAT manifest on the airplane but that he was aware that such a

document existed and was available. The ARFF captain stated that the ARFF response to the

accident scene would have been similar if they had known that HAZMAT was on board the

airplane, with the exception that all responding firefighters would have worn full personal

protective equipment.

NTSB Aircraft Accident Report

1.16 Tests and Research

1.16.1 Airplane Performance Study and Simulator Evaluations

The NTSB conducted an airplane performance study to estimate the airplane’s position,

airspeed, altitude, and other parameters throughout the accident flight. Data sources for the study

included the airplane’s FDR, FAA air traffic control (ATC) radar data, and weather information

services.

The study extracted aerodynamic coefficients from the FDR data. In addition, ATR

used its ATR 42-320 engineering model to which accident flight-specific information was

applied (including FDR data and weather information) to estimate control forces, AOA, and ice

accretion levels during the accident flight. Further, the NTSB used an ATR 42-320 engineering

flight simulator (EFS) provided by ATR to conduct qualitative testing to assess pilot workload

and to evaluate the airplane’s handling qualities with a flap asymmetry and ice accretion. The

EFS could not reproduce the accident airplane’s drag levels from ice accretion without

introducing unrealistic lift and pitch, roll, and yaw moments;

as a result, the “Normal Icing—

Boots On” ice levels were used with the knowledge that the drag effect was underestimated.

Pilots from ATR, Empire Airlines, and the NTSB participated in the EFS qualitative testing.

1.16.1.1 Effects of Flap Asymmetry

The flap asymmetry involved 8° to 10° of extension of the left flap and no extension of

the right flap. The greater flap extension on the left wing resulted in more lift and drag on that

wing; as a result, the airplane banked to the right. According to the FDR, the autopilot system

countered the airplane’s right bank by introducing 6° of left aileron input, for which the autopilot

moved the control wheel about 20° to the left.

The FDR indicated that, at 0434:55 (about 30 seconds after the flap asymmetry

occurred), the first officer reduced engine power from about 58 percent torque to about

3 percent. The autopilot, which was coupled to the ILS 17R approach, responded by increasing

the pitch attitude of the airplane to maintain the glideslope for the approach. The airplane’s

airspeed decreased to the point that, at 0435:30, the stick shaker activated, which disconnected

the autopilot when the airplane was about 1,000 feet agl. The performance study showed that the

stick shaker triggered at about an AOA of 7.2°, a load factor of 1.08 G, and an airspeed of

125 kts.

The FDR indicated that, after the autopilot disconnected (and the first officer began to

manually fly the approach), the first officer increased engine power to about 70 percent torque,

and the airplane began to deviate high and to the right of the ILS 17R glidepath.

Control force

The study results are estimates that include performance calculation assumptions and consider the inherent

limitations of the input data sources.

Title 14 CFR Part 60, which contains the qualification performance standards for airplane EFS models,

specifies that the effects of airframe icing (including airspeed decay and change in simulator pitch attitude) can be

based on subjective tests; that is, an EFS’s presentation of these effects is not required to match actual airplane

performance data obtained from airplane icing testing, incidents, or accidents.

The glidepath, which is the ideal flightpath for an ILS approach, is formed by the intersection of localizer

plane (which provides horizontal course guidance along the extended runway centerline) and glideslope plane

(which provides vertical course guidance).

NTSB Aircraft Accident Report

estimates that ATR provided indicated that the first officer was applying about 35 to 40 lbs of

rudder pedal pressure. ATR estimated that the average control wheel force required to manually

counter the flap asymmetry without retrimming the airplane was about 13 lbs.

At 0435:50, the captain took control of the airplane, and the estimated pilot forces exerted

on the cockpit controls peaked about 10 seconds later when the captain pulled back on the

control column, input LWD control wheel, and pushed on the left rudder pedal. ATR estimated

that the maximum forces exerted on the control column, control wheel, and rudder pedals were

about 36, 29, and 112 lbs, respectively. The FDR indicated that, at the time that these maximum

forces were applied, the airplane was about 500 feet agl and more than 300 feet right of course

(in the cockpit, this position would have displayed as “two dots” right on the course deviation

indicator),

and the stick shaker activated again. The performance study showed that the stick

shaker triggered at about an AOA of 6.8°, a load factor of 1.64 G, and an airspeed of 155 kts.

FDR data indicated that, at 0436:02 (about 2 seconds after the captain’s application of

high control forces), the control wheel and the ailerons moved from the deflected positions that

compensated for the flap asymmetry to their nominal positions. This movement was consistent

with the left-wing flaps retracting to approximately 4.5°, and the right-wing flaps extending to

approximately 4.5°, rendering all four wing flap panels symmetric, thus removing the need for

control inputs to counteract the flap asymmetry.

1.16.1.2 Effects of Airframe Ice Accretion

FAA inspectors who observed the wreckage about 3 hours 20 minutes after the accident

noted that the separated outboard section of the right aileron (which was found outside the fire

area) had an 8-inch patch of 1/8-inch thick ice adhered to the underside. The adhered ice patch

had jagged edges, as if some portions of ice had broken away from it. As noted in the

Aeronautical Information Manual (AIM),

section 7-1-21, “the effects of ice on aircraft are

cumulative; thrust is reduced, drag increases, lift lessens, and weight increases. The results are an

increase in stall speed and a deterioration of aircraft performance.”

The NTSB’s aerodynamic coefficient extraction and calculations made by ATR indicated

that airframe ice accretion during portions of the flight degraded the airplane’s performance. The

ATR calculations indicated that the airplane first encountered icing at the beginning of cruise

flight at flight level (FL) 180 (18,000 feet pressure altitude). ATR calculated that, at the end of

cruise flight at FL 180, a drag increase corresponding with about 15 percent of total power was

present. The aerodynamic coefficient extraction showed an increment in drag beyond the level

expected for an uncontaminated airframe from about 0335 to 0405 while the aircraft was in

cruise flight.

The ATR calculations indicated that, during the descent between 14,000 to about

5,000 feet msl, the airplane’s performance was consistent with what would be expected for an

Course sensitivity at this point in the approach is such that one dot represents about 150 feet of horizontal

course deviation.

Information obtained from the FAA’s website <http://www.faa.gov/air_traffic/publications/ATpubs/AIM/>

(accessed March 21, 2011).

NTSB Aircraft Accident Report

uncontaminated airframe (free from ice accumulation). The aerodynamic coefficient extraction

showed the drag level expected for an uncontaminated airframe from about 0405 to about 0432.

However, beginning about 0432 when the aircraft was at 5000 feet msl, an increment in drag

beyond the level expected for an uncontaminated airframe was again seen. ATR calculated that,

by the time that the flap asymmetry occurred, a drag increase corresponding with about

23 percent of total power was present. The EFS qualitative testing revealed that, following the

autopilot disconnect, the workload for the accident flight crew was high because of the flap

asymmetry, a 10-kt tailwind, and ice accretion. The testing showed that, when the approach was

continued without retrimming the airplane and at a speed lower than the minimum safe airspeed

of 143 KIAS (conditions similar to the accident flight), the controllability of the simulator was

often similar to what the FDR data showed for the accident airplane. However, when the

“AILERON MISTRIM” checklist in the Empire Airlines’ ATR quick reference handbook

(QRH) was applied

and engine power was set to maintain the minimum safe airspeed, the

pilots who conducted the EFS qualitative testing performed a missed approach (go-around

procedure when under IFR), completing the QRH procedures for the flap anomaly and the

reduced-flaps landing, and successfully landing the airplane.

FDR data indicated that, at 0436:18 (about 9 seconds before impact), the captain began a

pull back on the control column that lasted for several seconds. When the pull back began, the

airplane was at an airspeed of about 125 kts, descending through 230 feet agl about 1,500 feet

per minute (fpm); the pitch attitude was about -0.5°, and the load factor was about 0.9 Gs. In

response to the captain’s pull on the control column, the pitch attitude, the AOA, and the load

factor all began to increase. The performance study showed that when the stick shaker triggered

for the third time at 0436:19, the AOA was about 7°, the load factor was about 1.02 G, and the

airspeed was about 124 kts.

At 0436:20 (about 7 seconds before impact), with the AOA and the pitch attitude still

increasing, the load factor stopped increasing about 1.12 Gs, indicating that an aerodynamic stall

was imminent.

1.16.2 Terrain Awareness and Warning System

The airplane was equipped with a TAWS

designed to provide terrain caution alerts,

“pull up” warnings, and “avoid terrain” warnings. Data analysis showed that, when the TAWS

provided the flight crew with the aural “pull up” warnings, the airplane was about 1.5 miles from

the runway threshold at an altitude of 488 feet agl (as detected by the radio altimeter) with a

vertical descent rate of 2,050 fpm. An event analysis report provided by Aviation

Communication & Surveillance Systems, the manufacturer of the integrated system with the

TAWS feature, indicated that, during the accident approach, the airplane’s low radio altitude

combined with the rapid change in vertical speed resulted in the immediate TAWS warning

without a preceding caution alert.

For information about the procedure, which involves retrimming the airplane and then reengaging the

autopilot, see section 1.17.1.5.

This particular TAWS was part of an integrated system that also provided traffic alert and collision avoidance

functions.

NTSB Aircraft Accident Report

1.16.3 Flap System Component Examinations

The cockpit flap control lever was found in the 15° position. Examination of the flap

system components at the accident site found that the left wing and flap components were intact

and covered with soot. The right inboard and outboard flaps and the right inboard flap actuator

were extensively damaged by impact and fire. Figure 6 is the right side of the fuselage, showing

extensive fire damage to the right wing and flap components. The right outboard flap actuator

was found in the retracted position on the ground under the right wing. Damage to the right flaps

limited the extent to which preimpact system integrity could be determined. Examination of the

available components revealed no evidence of mechanical restriction of movement of the right

flaps.

Figure 6. Right side of fuselage showing extensive fire damage to right wing and flap

components.

Under normal circumstances, hydraulic pressure holds the flaps in the selected position.

The accident airplane’s hydraulic system was damaged by impact and fire, and much of the

hydraulic fluid routed to the flap actuators was lost because of the damage. Testing of fluid

samples obtained from all actuators (either from the actuator itself or the line leading to the

actuator) found that the liquid was consistent with hydraulic fluid.

X-ray examinations of all four flap actuators showed that their internal components were

in place. The left inboard and left outboard flap actuators functioned when tested on the test

bench at the manufacturer’s facility. Subsequent disassembly and examination of the left inboard

and outboard flap actuators revealed no anomalies. Several of the seals for the right inboard and

right outboard flap actuators were damaged from fire exposure, and the damage prevented the

NTSB Aircraft Accident Report

actuators from functioning on the test bench. Disassembly of the right inboard and outboard flap

actuators revealed no evidence of jamming.

Regarding the possibility that ice formation on the flap surfaces could have resulted in a

flap asymmetry, ATR stated that, in all types of icing conditions, runback icing does not exceed

15 to 20 percent of the wing chord on the upper wing surface and 30 to 40 percent of the wing

chord on the lower wing surface and that these limits are upstream of the flaps. Further, ATR

noted that ATR models having a similar wing/flap design have accrued 17,567,000 flight cycles

with no reports of flap asymmetry or jamming due to flight in icing conditions.

1.16.3.1 Hydraulic Fluid Testing

The piston rod for the right outboard flap actuator was found coated with a green material

for which chemical testing was inconclusive. Solid particles, which testing found to be consistent

with heated hydraulic fluid, were found in the hydraulic fluid lines and the connector port for the

right outboard flap actuator. No similar particles were found in any fluid samples from the

airplane’s hydraulic reservoir or from the examined hydraulic system filters.

1.16.3.2 Maintenance History

According to ATR, flap actuators are not subject to time limitations; they are serviced or

replaced based on their condition. At the time of the accident, the right inboard flap actuator and

the left outboard flap actuator were original equipment to the airplane (installed when it was

manufactured). The right outboard flap actuator and the left inboard flap actuator were installed

in February 2001.

Empire Airlines’ maintenance records indicated that a flap actuator hardware inspection

was completed on June 4, 2007. On October 31, 2007, during a takeoff configuration test

performed during taxi, the flaps did not extend but then functioned normally after subsequent

maintenance action to clean deicing fluid from the cannon plug. The airplane had been involved

in a January 21, 1998, accident in which a fuel-fed fire erupted in the right engine bay after

landing, which substantially damaged the airplane and exposed several components on the right

wing, including the flaps, to fire damage.

The airplane’s log books showed that the airplane

had been repaired, inspected, and returned to service.

1.16.3.3 Flap Asymmetry Event Involving Another Airplane

According to ATR, one previous flap asymmetry event was reported in 1996 involving an

ATR 42-500. The event occurred during a maintenance positioning flight (the maintenance was

related to a problem with the right powerplant that occurred on the previous day). During the

positioning flight, the asymmetry occurred when the flight crew commanded flap extension. The

airplane landed without incident, and the reason for the flap asymmetry was not determined.

The NTSB determined that the probable cause of the accident was “the improper overhaul of lug holes on the

fuel/oil heat exchanger. A factor was the lack of direction contained in the manufacturer’s overhaul manual for

working with the lug holes.” The report for this accident, NTSB case number NYC98FA062, is available online at

<http://www.ntsb.gov/aviationquery/index.aspx>.

NTSB Aircraft Accident Report

1.16.4 Deicing and Anti-Icing System Examinations

The pneumatic deicing boots on the left wing and the empennage were tested, and no

preimpact anomalies were identified. Impact and fire damage to the pneumatic boots on the right

wing prevented testing. The distributor valves for the airframe pneumatic boots were tested, and

all of the valves opened and closed as designed. All electrically heated components were

examined, and a resistance was obtained. Examination of the AAS annunciator panel by the

NTSB’s materials laboratory revealed no evidence of filament stretching in any bulbs.

1.16.5 Autopilot Disconnect Alert

The crew alerting computer, flight guidance computer, flight guidance controller, and

ADU were tested after the accident, and the autopilot disconnect and alerting systems functioned

normally. The autopilot disconnect aural alert sounded as designed. According to ATR, the

control wheel is equipped with an autopilot disconnect push button. Pressing the button while the

autopilot is engaged disconnects the autopilot. Pressing the button while the autopilot is

disconnected (whether disconnected manually with the button or by stick-shaker activation)

clears any messages from the ADU and silences any pending autopilot disconnect alert.

1.17 Organizational and Management Information

Empire Airlines, headquartered in Hayden, Idaho, provides cargo operations under

14 CFR Parts 135 and 121. The company, which has evolved under different names and has

provided a variety of services (including some passenger-carrying operations) since 1977,

began Part 121 operations in 1989. During that same year, Empire Airlines began a contract with

FedEx for cargo services using Fokker F27 airplanes. The company continued to evolve its fleet

and services and ended all passenger-carrying operations in 1995 to focus on cargo operations,

aircraft maintenance, and airline startups.

The company obtained a 14 CFR Part 145 repair

station certificate under the name of Empire Aerospace in 2001. It acquired its first ATR 42

airplane in 2003 and performed its first ATR 42 revenue flight for FedEx in 2004.

At the time of the accident, Empire Airlines had 16 bases of operation and served

51 destinations as a feeder operator for FedEx. According to the managing director of FedEx

feeder operations, FedEx’s relationship with its feeder operators is a fee-basis system in which

The company was first established as Clearwater Flying Service in 1977 and, by 1979, operated flights for

fire patrol, air pollution monitoring, air ambulance, charter, flight instruction, and transporting outfitters to remote

areas. In 1980, after acquiring another company, Clearwater Flying Service changed its name to Empire Airways

and expanded its services to include aircraft sales and maintenance. In 1986, the company contracted to operate

shuttle flights between ski resorts, and, in 1987, the company expanded with contracts to fly corporate employees

between offices in California and to perform flights in Alaska for the Naval Arctic Research Laboratory. Later that

same year, Empire Airways contracted with FedEx to fly and maintain Cessna 208 airplanes and to service routes

throughout the Pacific Northwest. In 1989, Empire Airways purchased Pacific Alaska and two Fokker F27 airplanes,

and the company changed its name to Empire Airlines.

Empire Airlines added turbojet airplanes for some operations in 1992. The company also assisted a startup

airline in Hawaii in 1992 and assisted another airline in Europe (to provide services for FedEx) in 1998. In 2006, the

company’s operations specifications authorized it to perform heavy maintenance on ATR 72 and Fokker F27

airplanes, and it later obtained heavy maintenance authorizations for de Havilland DHC-8 and Q400-series

airplanes.

NTSB Aircraft Accident Report

the operators are paid a management fee to crew and maintain the airplanes. At the time of the

accident, Empire Airlines operated 10 ATR 42, 3 ATR 72, and 35 Cessna 208 airplanes, all of

which were owned by FedEx. Empire Airlines employed about 250 people, including 108 pilots,

34 to 40 mechanics, and 7 flight followers.

1.17.1 Pilot Standard Operating Procedures and Training

Empire Airlines’ general operations manual (GOM) provided flight crews with

information on the company’s standard operating procedures (SOPs). Flight crew procedures

specific to the airplane were found in the ATR 42 AFM, which was an FAA-approved manual

that contained the normal, abnormal, and emergency procedures. The Empire Airlines ATR 42

QRH listed the procedures in a cockpit checklist format.

Empire Airlines provided its pilots with company indoctrination training at company

headquarters. Empire Airlines’ operations specifications authorized that all other initial,

recurrent, upgrade, and transition ground training, simulator training, testing, and checks were to

be provided by FlightSafety International, Inc., in Houston, Texas.

1.17.1.1 Flap Anomalies

According to the QRH, one procedure covered multiple flap anomaly conditions,

including a jam, uncoupling, or asymmetry. The procedure, “FLAPS

JAM/UNCOUPLED/ASYM,” specified that the pilot should position the flap control lever to

correspond with the actual flap position and apply the “REDUCED FLAPS LANDING”

procedure. Figure 7 is the QRH, page 2.21, with “FLAPS JAM/UNCOUPLED/ASYM” and

“REDUCED FLAPS LANDING” procedures. Pilots are trained to perform a go-around

maneuver in response to a flap anomaly so that they will have time to perform these required

procedures.

NTSB Aircraft Accident Report

Figure 7. Quick reference handbook page 2.21 with “FLAPS JAM/UNCOUPLED/ASYM” and

“REDUCED FLAPS LANDING” procedures.

According to the GOM, section 6-19, “Completion of Non-Normal Cockpit Checklist,”

flight crews are instructed that, “in the event of inoperative or malfunctioning flight control

system, if the airplane is controllable, complete only the applicable checklist procedures and do

not attempt any corrective action beyond those specified.” The GOM, section 6-4, “Circuit

Breakers Resetting,” stated that “tripped circuit breakers should not be reset in flight unless, in

the judgment of the PIC, it is necessary for the safe completion of the flight.”

The QRH, page 0.02 (directly beneath the Table of Contents), states the following

general information under the heading, “Procedures Initiation”:

NTSB Aircraft Accident Report

… No actions will be taken…until the flight path is stabilized… . Before

performing a procedure, the crew must assess the situation as a whole, taking into

consideration the failures, when fully identified, and the flight constraints

imposed.

Page 0.02 of the QRH also states that the PF is responsible for the engine power levers,

flightpath and airspeed control, aircraft configuration, and navigation, and the PM is responsible

for completing the appropriate checklist items, the actions on the overhead panel, the positioning

of the condition levers (which relate to propeller control), and communications. Figure 8 shows

the general information on page 0.02 of the QRH.

Figure 8. General information on page 0.02 of quick reference handbook.

NTSB Aircraft Accident Report

Both the captain and the first officer received training on flap anomalies during ground

school and simulator training at FlightSafety. The flap system anomalies that were covered

during ground school training included flap jam, flap unlock, flap uncoupled, flap asymmetry,

and no flaps. Flap control circuit breakers were discussed. The ground training reviewed the

QRH for flap anomalies item by item. For a flap asymmetry, the training taught the pilots that

the airplane will try to roll, which requires the pilot to take the controls. During simulator

training, flap anomalies were covered as part of both initial and recurrent training. The type of

anomaly was given at the discretion of the instructor (the syllabus did not specify which anomaly

to present). Pilots were trained to look at the flap position indicator in the cockpit, and they were

also taught to look out the side windows at the flap position fairings to verify the flap

conditions.

The NTSB interviewed FlightSafety instructors who stated that, regardless of which

anomaly was presented and when, the response procedure was the same. The pilots were taught

to initiate a go-around maneuver and complete the QRH, which required them to rebug the

airspeeds for the flap setting that they had. During postaccident interviews, the FlightSafety

instructors, the FAA principal operations inspector (POI) for Empire Airlines, Empire Airlines’

chief pilot, and Empire Airlines’ director of operations indicated that they would have expected

the accident flight crew to use the QRH procedure in response to the flap anomaly. During

subsequent testimony at an NTSB public hearing,

Empire Airlines’ chief pilot and director of

operations both stated that they would have expected the flight crewmembers to first assess the

situation and then reference the QRH after identifing the problem.

FlightSafety did not train pilots to troubleshoot a flap anomaly by checking the circuit

breakers, and none of the instructors who were interviewed had ever observed a trainee attempt

to do so in the flight simulator. One instructor could not recall any pilots ever attempting to land

with a flap abnormality during training and stated that it was “drilled into their heads to go

around” if a flap problem occurred.

During postaccident interviews and public hearing testimony, the captain stated that he

checked the cockpit flap position indicator and determined that the flaps had not extended. He

also stated that he attempted to troubleshoot the flap problem by checking the circuit breakers

because he “had to start somewhere” and that he needed to determine what the malfunction was

and then handle it with the checklist. The captain stated that he did not follow the abnormal

procedures checklist for the flaps because the airplane was on a short final approach, things were

happening quickly, and he “was too busy and did not know which checklist to run first or the

speeds.”

During postaccident interviews and public hearing testimony, the first officer stated that

she did not know there was a flap asymmetry and thought they had no flaps. She further stated

that she was “a little bit” concerned about the captain’s use of a flashlight to look at the circuit

breakers during the approach. She also stated that she thought he should have been more focused

on landing the airplane. The first officer stated that, had they executed a missed approach, they

Verifying the flap conditions was simulated in training as there were no actual flap position fairings in the

simulator.

For more information about the public hearing, see Appendix A.

NTSB Aircraft Accident Report

would have had time to use the QRH, get situated, and set their airspeeds. The first officer stated

that, after the captain decided that the approach would be continued, there was not enough time

to use the QRH.

1.17.1.2 Stabilized Approach and Go Around

Empire Airlines’ ATR 42 pilot handbook and flight training manual (FTM) both

addressed stabilized approach criteria and approach standardization procedures. According to the

FTM, approaches conducted in IMC should be stabilized by 1,000 feet in height above

touchdown, and approaches conducted in visual meteorological conditions (VMC) should be

stabilized by 500 feet in height above touchdown.

The FTM listed several criteria for the flight crew to determine that an approach would

be considered stabilized. These criteria included (but were not limited to) ensuring that only

small changes in heading and pitch are needed to maintain the correct track, the rate of descent

deviates no more than plus or minus 300 fpm from the target descent rate, the airplane’s airspeed

is not less than the required approach speed, the maximum sink rate does not exceed 1,000 fpm

(a special briefing should be performed for any approach requiring a greater sink rate), the power

setting is appropriate for the configuration, and all briefings and checklists had been performed.

According to the ATR 42 pilot handbook, “Flight Profiles & Briefings,” the “Descent and

Approach Awareness Procedures” indicated that, during an ILS approach, the PM is to “call out

any deviations from normal altitude, airspeed, or descent rates throughout the approach.” The

procedures noted that the “captain calls for ‘missed approach,’ if necessary.” The pilot

handbook, section 2-02-04, stated that, if a go-around maneuver is required, the PF should

initiate the procedure by calling out “go around,” pressing the “GA” (go around) button on the

engine power levers, advancing the engine power levers toward the white mark, and calling out

“max power, flaps 15.”

FlightSafety instructors who were interviewed stated that stick-shaker activation meets

the criteria for an unstabilized approach. During postaccident interviews, the captain stated that

he was trained to go around if a problem occurs during approach and that his decision to

continue the approach was based on the runway conditions,

icing conditions, and the flap

problem. The captain stated that he “made a judgment call not to go around because things

started piling up, and it was better to land than to go around.” He stated that he “just wanted to

land as soon as possible.”

During a postaccident interview, the first officer stated that stick-shaker activation

“would be a go-around situation” and that she asked the captain about going around because the

company policy following stick shaker was to perform a go-around maneuver. The first officer

According to the CVR transcript, the LBB approach controller advised the flight crew that the Mu readings

(friction coefficient values taken by runway friction-measuring equipment) for runway 35L at touchdown, midpoint,

and rollout were .24, .25, and .23, respectively. The captain acknowledged and then advised the first officer that,

during landing, she should keep the airplane on the runway centerline and not touch the brakes. Empire Airlines’

GOM (section 6-12, revision 43) provided a table that “converts braking action reports to friction coefficient values”

and showed that Mu readings between .19 and .34 corresponded with a “poor” braking action report. The GOM also

stated that “there is no direct correlation between friction coefficient values and braking action” and that flight crews

should use the values “together with other knowledge, including recent braking action reports and wind conditions.”

NTSB Aircraft Accident Report

indicated that asking was her way of saying that she wanted to go around “without stepping on

toes.” The first officer further stated that she did not challenge the captain’s decision to continue

because she felt that he had a good reason for not wanting to go around and that she trusted that

he was making the right decisions. The first officer also stated that, after the captain took control

of the airplane, she thought that she should have suggested again that they go around and that she

did not know why she did not say anything.

1.17.1.3 Terrain Awareness and Warning System

According to the GOM, if the flight crew receives a TAWS caution alert or warning in

conditions other than daylight VMC (in which the flight crew can positively confirm that a

collision with the terrain will not take place), the flight crew is to “react immediately.” The

procedures for responding to a “pull up” warning stated that the flight crew is to advance to

go-around power, disconnect the autopilot, level the wings while simultaneously executing a

positive pull up, set flaps to the go-around position, retract the landing gear, and maintain the

specified minimum airspeed

until terrain clearance is assured.

1.17.1.4 Airspeed Bugs

As described in section 1.6.4.1, Empire Airlines’ ATR 42 pilot handbook stated that, for

an approach in icing conditions, the flight crew should set the internal, yellow, white, and red

bugs to correspond with the V

mHB

30 icing, V

, V

mLB

0 normal, and V

mLB

0 icing airspeeds,

respectively. Flight crews were provided with an ATR 42 V-speed card pack that provided the

applicable minimum approach airspeeds based on the airplane’s weight. The same card that

provided approach airspeeds for each referenced airplane weight also provided airspeeds for use

during takeoff. The guidance provided to flight crews on setting the airspeed bugs did not

include any crosscheck procedures for verifying the information.

Postaccident examination of the cockpit found that the V-speed card pack was open to the

reference card for an airplane at 33,000 lbs gross weight. This card showed that, for this airplane

gross weight in icing conditions, V

mHB

30 was 116 kts, V

mLB

0 normal was 123 kts, and V

mLB

icing was 143 kts. V

was not depicted on the card. According to the ATR 42 pilot handbook,

is computed based on the normal approach airspeed (by either adding 5 kts to V

mHB

30 or

using 1.1 V

mca

, whichever is greater). Based on the card information, the approach airspeeds for

setting the internal, yellow, white, and red bugs were 116, 121, 123, and 143 kts, respectively,

and the takeoff airspeeds for setting the internal, yellow, white, and red bugs were 106, 112, 123,

and 143 kts, respectively.

According to the CVR transcript, during the captain’s approach briefing, he referenced

the “three thirty card” and read “one oh six [106 kts] is the icing speed…one oh six [106 kts] one

twelve [112 kts]…uh twenty three [123 kts] and forty three [143 kts].” Postaccident examination

of the captain’s and first officer’s airspeed indicators found that the captain’s internal, yellow,

white, and red bugs were set to 109, 110, 124, and 145 kts, respectively, and that the first

officer’s were set to 106, 112, 126, and 144 kts, respectively.

NTSB Aircraft Accident Report

1.17.1.5 Aileron Mistrim with Autopilot Engaged

With the autopilot system engaged, the roll anomaly induced by a flap asymmetry will

generate “RETRIM ROLL R [or L] WING DN” and “AILERON MISTRIM” messages on the

ADU. According to the manufacturer, a flight crew’s response to these messages should be to

hold the control wheel firmly, disengage the autopilot, and adjust the lateral trims. Upon

completion of these items, the autopilot may be reengaged. The AFM stated that, since the

autopilot may mask tactile cues that indicate adverse changes in airplane handling

characteristics, use of the autopilot is prohibited when unusual lateral trim requirements or

autopilot trim warnings are encountered when the airplane is in icing conditions.

During

postaccident interviews, both the captain and the first officer indicated that they did not see any

messages on the ADU.

1.17.1.6 Deicing and Anti-Icing Systems

The AFM defined three levels of icing equipment use and stated that level 3 protection,

which the flight crew used during the accident descent, should be used any time that ice is

observed accumulating on the airframe. According to the AFM, level 3 icing protection includes

all level 1 and 2 items, plus the use of continuous relight engine ignition and the airframe

pneumatic deicing boots.

Flight crew selection of the flight control horn anti-icing system

(used in level 2 and 3 ice protection) lowers the AOA stall warning threshold and triggers the

illumination of the green ICING AOA annunciator. Once airframe pneumatic boot use is selected

by the flight crew (by pushing the “AIRFRAME” button), the boot system inflates in sequence

continuously without further crew input.

According to the FCOM, the AAS’s amber ICING annunciator illuminates as soon as

(and as long as) the system’s ice detector senses that ice is building up. The flight crew was

trained to respond to AAS annunciators in the cockpit by increasing the minimum maneuvering

and operating airspeeds and selecting the anti-icing system when flying in icing conditions,

selecting the deicing system at the first indication of airframe ice accumulation, and switching

the deicing system off once ice is no longer present on the airframe. Once the ice detector no

longer senses that ice is accumulating, the system triggers the blue DEICING annunciator to

flash, alerting the flight crew to visually check for the absence of ice and determine if the deicing

system should be turned off.

The FCOM cautions that freezing rain is composed of SLD that may transform into clear

ice when it impacts the airplane’s skin in below-freezing temperatures. The FCOM states that

these large droplets may not freeze immediately upon impact and that the ice detector may

become “inefficient” (ineffective). According to ATR’s Cold Weather Operations publication,

The AFM also stated that use of the autopilot is prohibited any time that severe icing exists.

Level 1 protection, which was to be used during all flights, included use of the probe heat and windshield

heat. Level 2 protection, which was to be used for all takeoffs and all flight operations in atmospheric icing

conditions, included the level 1 items plus the use of propeller, flight control horn, and side window heat; engine

inertial separators and pneumatic boots; minimum propeller speed of 86 percent; illumination of the AOA light; and

icing speeds bugged and observed.

ATR Flight Operations Services, Cold Weather Operations: Be Prepared for Icing (2008). This publication

was updated in March 2011, and the current edition can be obtained from ATR’s website

NTSB Aircraft Accident Report

flight crews can visually assess ice accumulation by observing the propeller spinners,

windshield, airframe leading edges, wipers, and side windows.

According to the AFM limitations, section 4-5, if severe icing conditions are encountered

in flight, the flight crew’s procedures include escaping the severe icing conditions and notifying

ATC

According to the AFM,

severe icing may result from environmental conditions outside of those for which

the airplane is certificated. Flight in freezing rain, freezing drizzle, or mixed icing

conditions (supercooled liquid water and ice crystals) may result in ice buildup on

protected surfaces exceeding the capabilities of the ice protection system or may

result in ice forming aft of the protected surfaces.

The AFM states that the visual cues to identify severe icing include “ice covering all or a

substantial part of the unheated portion of either side window,” and/or an “unexpected decrease

in speed or rate of climb,” and/or the following secondary indications:

• Water splashing and streaming on the windshield.

• Unusually extensive ice accreted on the airframe in areas not normally observed

to collect ice.

• Accumulation of ice on the lower surface of the wing aft of the protected area.

• Accumulation of ice on the propeller spinner farther aft than normally observed.

In a postaccident interview, the captain stated that the airplane had encountered

“moderate bordering on severe” rime ice at FL 180. He stated that he selected level 3 ice

protection and that he noticed that, during the icing encounter, the airplane was cruising at

180 KIAS, whereas it would typically cruise at 200 to 210 KIAS. He stated that he requested a

descent to 14,000 feet msl and that, during the descent, “substantial amounts of ice” came off the

airplane. He noted that some ice from that encounter was still on the airframe when the flight

started its descent to LBB. Further, in a postaccident interview, the first officer stated that the

airplane began to accumulate ice again as it descended through 6,000 feet msl. She recalled that

level 3 ice protection had been activated and that she could see ice adhering on the propeller

spinner.

1.17.1.7 Crew Resource Management

CRM is the effective use of all available resources, people, information, and equipment to

achieve safe and efficient flight operations.

CRM involves crewmembers incorporating and

synchronizing the tasks required of them in a correct and timely manner

and communicating

<http://www.atraircraft.com/content/media/downloads/coldweatheroperations_2011_20.pdf> (accessed April 14,

2011).

J. Lauber, “Resource Management in the Cockpit,” Air Line Pilot, vol. 53 (1984), pp. 20-23.

J.A. Cannon-Bowers Volpe and others, “Defining Team Competencies and Establishing Team Training

Requirements,” in R. Guzzo, E. Salas, & Associates, eds., Team Effectiveness and Decision Making in

Organizations (San Francisco: Jossey-Bass, 1995), pp. 333-380.

NTSB Aircraft Accident Report

and coordinating effectively. FlightSafety provided CRM training to Empire Airlines’

crewmembers during initial training, and CRM was reviewed during recurrent training.

The CRM training provided during the pilots’ initial training consisted of 2 days

(16 hours) in the classroom and included lectures, case study reviews, and breakout and team

building sessions with crewmembers. Lecture topics included command and leadership,

situational awareness, communication, decision making, error management, teamwork and

synergy, fatigue, risk assessment, and automation management. The CRM training materials

included two presentation slides on assertion that encouraged pilots to speak up with

“appropriate persistence until there is a clear resolution.” The FlightSafety ATR program

manager stated that Empire Airlines’ CRM training program did not include role-playing

exercises in the classroom training. During a postaccident interview, the first officer stated that

she could not recall what was presented during the 2-day CRM training course, including

whether or not assertiveness was addressed.

According to Empire Airlines’ FTM, the CRM review provided during recurrent training

was included within the “Crew Duties and Responsibilities” module and in “Flight Module 1.”

According to Empire Airlines’ FTM, section 8-1-6, applied CRM is monitored and practiced in

each flight simulator module. One FlightSafety instructor stated that CRM was evaluated

throughout the simulator training because the ATR 42 is a two-pilot airplane, and he stated that

he would point out CRM issues in real time if he felt that it was necessary. Another instructor

indicated that instructors had the tools and techniques to teach CRM during the prebriefing, in

the simulator, and during the debriefing.

1.17.1.8 Weather Information and Training

Empire Airlines’ chief pilot indicated that, during basic indoctrination training, pilots

reviewed Empire Airlines’ aircraft deicing program (ADP); the FAA’s meteorology handbook;

the National Aeronautics and Space Administration’s (NASA) video, Tailplane Icing;

operations in icing for corporate aircraft video; and handouts of flight releases to familiarize

pilots with the format.

Empire Airlines’ chief pilot also indicated his belief that, during training at FlightSafety,

pilots reviewed the ATR Cold Weather Operations publication

and discussed meteorological

conditions likely to cause freezing drizzle and freezing rain. Interviews with FlightSafety

instructors revealed that several were familiar with the ATR publication but that none of the

instructors provided or discussed it during training. A review of Empire Airlines’ FTM revealed

that it included a task-listing document that indicated that Empire Airlines (not FlightSafety) was

The NASA video describes the tailplane stall condition that can result from horizontal stabilizer icing, its

warning signs, and the suggested recovery procedures. The video’s recovery procedures apply to airplanes that have

unpowered controls, that is, airplanes with horizontal stabilizers that have a fixed leading edge and movable elevator

surfaces that rely on aerodynamic balance to keep the control forces neutral. The ATR 42 is equipped with such

unpowered controls (also known as reversible controls). All airplanes that are certificated under 14 CFR Part 25

have been tested for tailplane stall susceptibility under at least the Part 25, Appendix C, icing conditions and have

demonstrated that they can be safely operated in such conditions. For conditions that are outside the Part 25,

Appendix C, icing envelope, the performance and handling characteristics of the airplanes are not required to have

been evaluated.

Cold Weather Operations: Be Prepared for Icing (2008).

NTSB Aircraft Accident Report

to provide the required training as part of the meteorology module, but Empire Airlines did not

provide it.

A review of the ATR Cold Weather Operations publication

revealed that it provided

guidance that advised pilots that icing is a major concern for commuter aircraft. The publication

stated that, due to their typical operating altitudes and airspeeds, turboprop aircraft fly where

icing conditions are most likely to occur. The publication also provided information about icing

meteorological phenomena, aircraft systems available to prevent and control ice accumulation,

aircraft performance loss due to ice contamination on the aerodynamic surfaces of the airplane,

and the procedures to apply on the ground and during flight when faced with icing conditions.

1.17.1.9 Stall and Near-Stall Recoveries

A review of Empire Airlines’ FTM revealed that company pilots were taught that at first

stall indication,

they should set maximum power, reduce pitch, roll wings level, and then,

during stall recovery, return to the entry altitude.

In June 2010, a stall recovery procedure was added to the ATR 42 AFM. The procedure,

“RECOVERY AFTER STALL OR UNCOMMANDED ROLL CONTROL,” specifies that the

pilot should push firmly on the control wheel and then set maximum power.

1.17.2 Aircraft Deicing Program (Ground)

In accordance with 14 CFR 121.629, Empire Airlines had an FAA-approved ADP to

address procedures for ensuring that no airplanes are dispatched or released for takeoff with

unsafe frozen contaminants on their surfaces. The program conformed to the guidance in

AC 120-60B, “Ground Deicing and Anti-Icing Program,” which provides information on

industry standards and practices related to ground deicing and anti-icing programs. The ADP

contained holdover tables that provided guidelines on the amount of time that deicing and

anti-icing fluid would protect an airplane’s critical surfaces from frozen contaminants, such as

active frost, freezing fog, snow, snow grains, freezing drizzle, light freezing rain, and rain on

cold-soaked wings.

1.17.3 Flight Dispatch

Empire Airlines’ flight dispatching functions were performed from the communications

center at its headquarters. As a 14 CFR Part 121 supplemental air carrier, Empire Airlines was

authorized to use flight followers (rather than licensed dispatchers) to authorize each flight

release, provide the necessary preflight weather products, and perform flight following of the

flight to the destination. All of Empire Airlines’ flight followers (with the exception of one who

Cold Weather Operations: Be Prepared for Icing (2008).

As defined by Empire Airlines, the first indication of a stall is the recognition by the pilot of the stall warning

aural “cricket,” the stick shaker, the stick pusher, or buffeting/lateral instability.

NTSB Aircraft Accident Report

was newly hired) were licensed dispatchers.

Flight followers communicated with flight crews

through the use of telephones, fax machines, and air-to-ground communications services.

The flight release for the accident flight was issued about 0211 and listed Roswell

International Air Center Airport, Roswell, New Mexico (located about 137 miles west of LBB),

as the alternate airport. The weather document indicated that the aviation routine weather reports

for LBB were reporting IFR conditions with light freezing drizzle and mist and that the aviation

terminal forecast expected the IFR conditions and light freezing drizzle to continue beyond the

estimated time of arrival. Several observations surrounding LBB also reported mixed freezing

precipitation across the area and indicated that these were not isolated events. The document also

contained the full AIRMET Zulu series (which pertains to icing) current over the area for

occasional moderate icing conditions below 22,000 feet msl. The document indicated that the

forecast for the Roswell International Air Center Airport expected wind from the north-northwest

at 12 kts, visibility better than 6 miles, no precipitation, and a ceiling broken at 1,600 feet agl.

According to the GOM, section 8-9, “when light freezing rain, light or moderate freezing

drizzle, or light, moderate or heavy snow is falling, aircraft may land.” The GOM also stated

that, “when light freezing rain, light to moderate freezing drizzle, or light, moderate snow is

falling, [an ATR] aircraft may take off, provided it is prepared in accordance with approved

deicing procedures.”

Empire Airlines’ director of operations stated that the airline developed

this guidance using the titles from the ADP holdover tables after receiving flight crew requests

for guidance related to operating in icing conditions.

During a postaccident interview, the flight follower who initiated preparations of the

flight release for the MAF-to-LBB flight leg indicated that she was aware of the freezing drizzle

conditions surrounding LBB. Another flight follower released the flight. The flight followers

who handled the accident flight’s release and the dispatch manager indicated that the GOM

contained no restriction that prevented ATR airplanes from operating in light freezing drizzle

and light freezing rain. All of them indicated that they believed that the ATR airplanes could

handle light freezing rain and freezing drizzle conditions but that the Cessna 208 airplanes

(which the flight followers indicated were prohibited from operating in those conditions) could

not. Both the dispatch manager and one of the flight followers indicated that as long as the

airplanes met the requirements for ground deicing, they could be released for flight.

During postaccident interviews, both the captain and the first officer stated that the

company did not pressure them to depart or arrive on time. The flight follower who released the

accident flight indicated that he had a “normal conversation” with the captain and that the

weather conditions were not that bad compared to other weather he had seen over the years. The

flight followers who were interviewed indicated that, if a captain disagreed with the flight

followers, the flight followers typically backed up the captain’s decision, and the captain had the

final say. During public hearing testimony, Empire Airlines’ manager of dispatch stated that if a

Although the flight followers were licensed dispatchers, their duties were defined more closely with the role

of flight follower and did not require the same initial and recurrent training that is required of a dispatcher.

The GOM prohibited the company’s Cessna 208 airplanes from departing, operating en route, or landing in

freezing rain or freezing drizzle of any intensity.

NTSB Aircraft Accident Report

captain was not comfortable launching due to weather, the flight follower would support the

captain’s decision.

A cursory review of other 14 CFR Part 121 and 135 operators’ icing guidance (which

included some operators that operated ATR or other turbine-powered, deice boot-equipped

airplanes) revealed that the guidance ranged from prohibitive to permissive. For example, some

guidance stated that “flight” or “takeoffs” in freezing rain and freezing drizzle was “not

approved” or “prohibited”; other guidance warned that “flight in freezing rain [or] freezing

drizzle…may result in ice buildup…exceeding the capability of the ice protection system” or that

“visible rain at temperatures below 0° C” may be conducive to “severe in-flight icing”; and other

guidance indicated that “aircraft may operate when light freezing rain [or] light to moderate

freezing drizzle…is falling…provided that it is prepared in accordance with approved deicing

procedures” or that “when light freezing rain [or] light or moderate freezing drizzle…is falling,

aircraft may land.”

1.17.4 Hazardous Materials

The GOM contains Empire Airlines’ procedures related to HAZMAT cargo. Specifically,

section 10-16 states that emergency response information (required in accordance with 49 CFR

172.600[c]) for use in mitigating an incident involving the HAZMAT cargo must be immediately

available where the HAZMAT is received, stored, or handled during transportation and must be

immediately available on board the aircraft. Section 10-17 indicates that, before departure, the

PIC must be provided written information about the HAZMAT cargo, proper shipping name,

hazard class, identification number, quantity, and location on board the airplane. The GOM

states that a copy of the HAZMAT document provided to the PIC is subject to the following (in

accordance with 49 CFR 175.33[c]):

This information must be readily accessible at the airport of departure and the

intended airport of arrival for the duration of the flight leg and [must be made…]

immediately available to any representative of federal, state, or local government

agency who is responding to an incident involving the flight.

1.17.5 Federal Aviation Administration Oversight

The FAA’s flight standards district office (FSDO) in Spokane, Washington, provided

oversight for Empire Airlines. The FSDO’s oversight of Empire Airlines’ 14 CFR Part 121

operations was managed through the FAA’s air transportation oversight system (ATOS).

According to the FAA, ATOS provides a means of oversight that allows air carriers to meet and

address all regulatory requirements, enables FAA inspectors to develop methodology for

continued operational safety, and provides FAA inspectors and air carriers the ability to manage

risk. The ATOS oversight requirements included oversight of the content of an airline’s manual

and the training of the flight crewmembers.

ATOS oversight requirements are contained in FAA Order 8900.1, Flight Standards Information

Management System, Volume 10, Chapter 1, Figure 10-7,“System/Subsystem/Element Chart—Operations and

Cabin Safety Elements.” Information obtained from the FAA’s website

<http://fsims.faa.gov/PICResults.aspx?mode=EBookContents> (accessed April 14, 2011).

NTSB Aircraft Accident Report

The FAA POI for Empire Airlines had been assigned to the company for more than

11 years. He stated that the GOM was the FAA-approved manual and that the pilot handbook

was “accepted.” He stated that he was the POI for the company when operations in freezing

drizzle were approved. He also stated that Empire Airlines had recently developed an internal

audit program for flight operations. The POI described Empire Airlines as a company with good

management people who self-report any problems.

The FAA principal maintenance inspector (PMI) for Empire Airlines had been assigned

to the company since 2000. He described Empire Airlines as very cooperative, very responsive in

correcting any identified problems, and consistent about submitting self-disclosures. The PMI

stated that his work program had been managed through ATOS since October 2006. He stated

that the ATOS process enabled him to develop the depth of inspection required.

1.17.6 Previous Accidents

The company had two previous fatal accidents in 1995 and 2000, both of which involved

single-pilot cargo flights that were operated under 14 CFR Part 135. On January 11, 1995, a

FedEx-owned Cessna 208 operated by Empire Airlines collided with a ridgeline in Pleasanton,

California, while it was cleared for a visual approach to the Metropolitan Oakland International

Airport, Oakland, California.

The NTSB determined that the probable cause of the accident

was the pilot’s failure to maintain both visual contact with terrain and sufficient altitude for

terrain clearance.

On October 9, 2000, a FedEx-owned Cessna 208 operated by Empire

Airlines collided with wooded terrain in Lummi Island, Washington, during a flight conducted in

IMC under a special visual flight rules (SVFR) clearance.

The NTSB determined that the

probable cause of the accident was the pilot’s attempted flight into known adverse weather

conditions and the pilot’s subsequent failure to maintain altitude above or clearance from the

trees.

The report for this accident, NTSB case number LAX95FA076, is available online at

<http://www.ntsb.gov/aviationquery/index.aspx>.

Contributing to the accident were the pilot’s decision to initiate a descent 14 miles from the airport and the

weather, specifically the cloud conditions and darkness.

The FAA defines SVFR conditions as meteorological conditions that are less than those required for basic

visual flight rules flight in controlled airspace. SVFR operations must be requested by the pilot and approved by

ATC. According to the accident report, the accident flight’s leg could not be conducted under IFR because there was

no published instrument approach procedure at the destination airport in Eastsound, Washington. The report for this

accident, NTSB case number SEA01FA001, is available online at <http://www.ntsb.gov/aviationquery/index.aspx>.

Contributing to the accident were the low ceilings, fog, the pilot’s low-altitude flight, rising terrain, and trees.

NTSB Aircraft Accident Report

1.18 Additional Information

1.18.1 Postaccident Safety Action

1.18.1.1 Federal Aviation Administration

Safety Alert for Operators: In-Flight Icing Operations and Training

On March 16, 2010, the FAA issued Safety Alert for Operators (SAFO) 10006, “In-Flight

Icing Operations and Training Recommendations,” to encourage all operators to review and

amend, if necessary, their flight crewmember and dispatcher training programs to ensure that

their programs address SLD icing conditions. The SAFO notes that SLD conditions, which the

FAA considers synonymous with freezing drizzle and freezing rain aloft, are “icing conditions

containing droplets larger than those required to be demonstrated in aircraft icing certification

criteria. SLD may result in ice formation beyond the capabilities of the airplane’s ice protection

system to provide adequate ice protection.”

The SAFO referenced the NTSB’s investigation of the October 31, 1994, accident

involving an ATR 72-212 that crashed following an in-flight encounter with icing in Roselawn,

Indiana.

In the SAFO, the FAA recommended that the directors of safety and directors of

operations for 14 CFR Part 121 and 135 operators and the training managers for all operators

ensure that their training programs include a review of the following: meteorological conditions

likely to cause SLD conditions; identification of weather information sources (such as AIRMET,

significant meteorological information [SIGMET], center weather advisories, hazardous in-flight

weather advisory service, current icing potential, and forecast icing potential) and their use as it

relates to flight crewmembers’ and dispatchers’ preflight planning and in-flight decision making;

discussion of company and ATC procedures for pilot reports on severe icing; discussion of flight

crewmembers’ identification of severe icing conditions associated with freezing rain and

freezing drizzle, including immediate exit procedures and ATC procedures should severe icing

be encountered; and a review of the FAA publication, “Roll Upset in Severe Icing.”

Proposed Rulemaking for Airplane Certification Requirements for Flight in Icing

Conditions

On June 29, 2010, in response to several NTSB safety recommendations (several of

which were issued during the NTSB’s investigation of the October 31, 1994, accident in

Roselawn, Indiana,

and many of which superseded or expanded upon previous safety

recommendations issued in 1981),

the FAA issued a notice of proposed rulemaking (NPRM),

See In-flight Icing Encounter and Loss of Control, Simmons Airlines, d.b.a. American Eagle Flight 4184,

Avions de Transport Regional (ATR) Model 72-212, N401AM, Roselawn, Indiana, October 31, 1994; Volume 1:

Safety Board Report (Revision 9/13/02), Aircraft Accident Report NTSB/AAR-96/01 (Washington, DC: National

Transportation Safety Board, 1996). <http://www.ntsb.gov/publictn/1996/AAR9601.pdf>.

“Roll Upset in Severe Icing,” Flight Safety Digest (January 1996). Information obtained from the Flight

Safety Foundation’s website <http://flightsafety.org/fsd/fsd_jan96.pdf> (accessed March 21, 2011).

NTSB/AAR-96/01.

The NTSB first identified the need for the FAA to review and revise the icing criteria in 14 CFR Part 25,

Appendix C, in a 1981 safety report. For more information, see Aircraft Icing Avoidance and Protection,

September 9, 1981, Safety Report NTSB/SR-81/01 (Washington, DC: National Transportation Safety Board, 1981).

NTSB Aircraft Accident Report

“Airplane and Engine Certification Requirements in Supercooled Large Drop, Mixed Phase, and

Ice Crystal Icing Conditions,” intended to improve the safety of transport-category airplanes

operating in SLD, mixed-phase, and ice-crystal icing conditions.

Included in the NPRM were

proposals for expanding the certification icing environment for transport-category airplanes to

include freezing rain and freezing drizzle and to require airplanes most affected by SLD icing

conditions to meet certain safety standards in the expanded certification icing environment,

including additional airplane performance and handling requirements. The NPRM noted that

freezing drizzle is outside the icing envelope

currently specified in 14 CFR Part 25,

Appendix C, which contains the certification requirements for transport-category airplanes.

1.18.1.2 Empire Airlines

On March 17, 2009, Empire Airlines’ Director of Flight Operations wrote an email to

NTSB investigators detailing the company’s postaccident actions. These actions included the

following:

Training Guidance for Flap Anomalies

Empire Airlines distributed guidance to all check airmen and the FlightSafety ATR

program manager that stated the following:

In the course of conducting flight training or checking in the ATR (airplane or

simulator) and specifically while executing flap malfunction and reduced flap

landing, do not let the crew think that you want them to continue the approach

without taking the time to complete the QRH procedure(s), reset the speed bugs,

rebrief the approach, etc…Depending on where the crew recognizes the flap

problem, they should probably ask ATC (instructor/check airman) for a delay

vector or hold while they complete the QRH before continuing the approach. The

point is we don’t want to train or check differently than we want the crew to fly

the aircraft. We should expect them to delay the approach until all the QRH

procedures and briefs are complete.

Flight Information Bulletin for Airspeed Bugs

On February 23, 2009, Empire Airlines issued Flight Information Bulletin 09-01 to all

ATR crewmembers explaining the proper determination and setting of airspeed bugs for flight in

icing conditions. The company also added expanded procedures to the ATR pilot handbook for

flight crews to be aware of errors in setting airspeed bugs and added a procedure for the setting

and readback of the bugs during the departure and approach briefings.

75 Federal Register (FR) 37311-37339 (June 29, 2010).

According to the NPRM, the term “icing envelope” refers to “the environmental icing conditions within

which the airplane must be shown to be able to safely operate.”

For information about the NTSB’s August 27, 2010, comments to the FAA regarding the NPRM, see

section 2.4.1.2.

NTSB Aircraft Accident Report

Flight Operations Bulletin for Icing

On February 27, 2009, Empire Airlines issued Flight Operations Bulletin 09-04, which

contained information to supersede the information contained in the GOM, section 8-9, “Icing.”

The flight operations bulletin, which was intended for both ATR flight crews and flight

followers, included the following information:

Freezing rain…and freezing drizzle…are not covered by the certification icing

envelope as defined in Part 25, Appendix C. Takeoff or landing operations in

known or reported [freezing rain or freezing drizzle] of any intensity, are

prohibited.

Special Emphasis Icing Training

Empire Airlines implemented special emphasis icing training, which increased the time

spent during ground training for all ATR pilots and dispatchers and included a review of the

ATR Cold Weather Operations publication,

the NASA Tailplane Icing video, and the 14 CFR

Part 25, Appendix C, icing certification envelope. The training also included a review of SLD;

severe icing; icing reports, forecasts, advisories, AIRMETs, and SIGMETs; and a review of

aircraft performance and the recommended maximum crosswind and tailwind components with

respect to operations with contaminated runways.

1.18.1.3 FedEx Feeder Operations

In April 2009, FedEx facilitated a safety summit with its feeder operators to address the

circumstances of this accident and facilitate training improvement and adherence to best

practices. As a result of the safety summit, the ATR training curriculum taught at FlightSafety

was modified to include stick-shaker and stick-pusher recognition and recovery training, special

purpose operational training, and line-oriented flight training (one module of which is still under

development). ATR initial training was also increased with two sessions added. In addition,

company instructors were placed on site at FlightSafety to perform company-specific training

and check rides to complement FlightSafety’s instruction.

All FedEx feeder operators developed “no-go” weather items (similar to those contained

in Empire Airlines’ Flight Operations Bulletin 09-04) that prohibit takeoff or landing operations

in known or reported freezing rain or freezing drizzle. All of the operators have implemented

procedures for ensuring that airspeed bugs are properly used and verified during all phases of

flight. Ice evidence probes (IEPs) have been installed on all ATR airplanes that were not so

equipped (the installations were completed by October 2009).

Cold Weather Operations: Be Prepared for Icing (2008).

NTSB Aircraft Accident Report

1.18.2 Previously Issued Safety Recommendations

1.18.2.1 Flight Crew Monitoring and Workload Management Skills for Single and

Multiple Abnormal Situations

On February 23, 2010, as a result of its investigation of the February 12, 2009, accident

involving a Colgan Air (operating as Continental Connection) Bombardier DHC-8-400, which

crashed during an instrument approach in Buffalo, New York,

the NTSB issued Safety

Recommendation A-10-10, which asked the FAA to do the following:

Require 14 Code of Federal Regulations Part 121, 135, and 91K operators to

review their standard operating procedures to verify that they are consistent with

the flight crew monitoring techniques described in Advisory Circular (AC)

120-71A, “Standard Operating Procedures for Flight Deck Crewmembers”; if the

procedures are found not to be consistent, revise the procedures according to the

AC guidance to promote effective monitoring.

On June 22, 2010, the FAA responded that it was considering options such as issuance of

a SAFO or an information for operators (InFO) bulletin to address this recommendation. On

January 25, 2011, the NTSB indicated that although such action may represent part of an

acceptable response to this recommendation, the FAA must also ensure that the guidance

provided in documents like SAFOs is implemented.

Pending the recommended actions, the

NTSB classified Safety Recommendation A-10-10 “Open—Acceptable Response.”

As a result of its investigation of the February 16, 2005, accident involving a Cessna

Citation 560 that crashed during an ILS approach in icing conditions,

the NTSB issued Safety

Recommendation A-07-13, which asked the FAA to do the following:

Require that all pilot training programs be modified to contain modules that teach

and emphasize monitoring skills and workload management and include

opportunities to practice and demonstrate proficiency in these areas.

In this accident, “the flight crew’s failure to monitor airspeed” was a contributing factor. See Loss of Control

on Approach, Colgan Air, Inc., Operating as Continental Connection Flight 3407, Bombardier DHC-8-400,

N200WQ, Clarence Center, New York, February 12, 2009, Aircraft Accident Report NTSB/AAR-10/01

(Washington, DC: National Transportation Safety Board, 2010).

<http://www.ntsb.gov/publictn/2010/AAR1001.pdf>.

At the same time Safety Recommendation A-10-10 was issued, the NTSB also issued Safety

Recommendation A-10-31, which concerns the need to ensure that guidance provided in documents like SAFOs has

been implemented. Safety Recommendation A-10-31 asked the FAA to do the following: “Implement a process to

document that all 14 Code of Federal Regulations Part 121, 135, and 91K operators have taken appropriate action in

response to safety-critical information transmitted through the safety alert for operators process or another method.”

On January 25, 2011, the NTSB classified this recommendation “Open—Acceptable Response,” pending FAA

action.

The NTSB determined that the probable cause of the accident included “the flight crew’s failure to

effectively monitor and maintain airspeed.” See Crash During Approach to Landing, Circuit City Stores, Inc.,

Cessna Citation 560, N500AT, Pueblo, Colorado, February 16, 2005, Aircraft Accident Report NTSB/AAR-07/02

(Washington, DC: National Transportation Safety Board, 2007).

<http://www.ntsb.gov/ntsb/publictn/2007/AAR0702.pdf>.

NTSB Aircraft Accident Report

On May 17, 2007, the FAA stated that monitoring skills and workload management are

part of CRM and that current CRM regulations adequately address the issues in this

recommendation. However, the FAA indicated that it would consider identifying in its work

program a list of required inspections, reemphasizing to the regional and FSDO managers the

need to validate the training that is already required and verify its effectiveness. On September 9,

2008, the NTSB responded that the FAA’s list should include a strong emphasis on monitoring

and workload management components of the CRM program. On February 23, 2010, as a result

of the FAA’s lack of progress in response to the safety recommendation and the NTSB’s

subsequent investigation of the Colgan Air accident,

the NTSB reiterated A-07-13 and

classified it “Open—Unacceptable Response.”

On May 19, 2009, as a result of the NTSB’s investigation of the September 28, 2007,

accident involving American Airlines flight 1400, a McDonnell Douglas DC-9-82, which

experienced an in-flight engine fire and nose landing gear anomaly that presented the flight crew

with multiple, simultaneous emergency and abnormal situations,

the NTSB issued Safety

Recommendations A-09-24 and -25, which asked the FAA to do the following:

Establish best practices for conducting both single and multiple emergency and

abnormal situations training. (A-09-24)

Once the best practices for both single and multiple emergency and abnormal

situations training asked for in Safety Recommendation A-09-24 have been

established, require that these best practices be incorporated into all operators’

approved training programs. (A-09-25)

On August 11, 2009, the FAA responded that it supports “realistic, scenario-based

training that serves to develop pilot psychomotor skills as well as critical thinking and

decision-making abilities in response to the range of unusual or unanticipated events that may

occur during flight operations.” The FAA stated that, on January 12, 2009, it published an

NPRM, “Qualification, Service, and Use of Crewmembers and Aircraft Dispatchers,”

that

included such scenario-based training for all 14 CFR Part 121 operators. The FAA also indicated

that it planned to revise FAA Order 8900.1, “Flight Standards Information Management

System,” to require POIs to withdraw approval of any Part 121 training program that does not

include such training.

On April 2, 2010, the NTSB responded that, although issuance of the final rule proposed

in the NPRM would give the FAA the regulatory framework to require the training, the NTSB

remained concerned that, without a guide to best practices, inspectors and operators would have

difficulty creating effective programs. Pending the outcome of the FAA’s planned review of

regulatory and policy framework to determine whether additional guidance or requirements are

NTSB/AAR-10/01.

See In-Flight Left Engine Fire, American Airlines Flight 1400, McDonnell Douglas DC-9-82, N454AA,

St. Louis, Missouri, September 28, 2007, Aircraft Accident Report NTSB/AAR-09/03 (Washington, DC: National

Transportation Safety Board, 2009). <http://www.ntsb.gov/publictn/2009/AAR0903.pdf>.

74 FR 1280 (January 12, 2009).

NTSB Aircraft Accident Report

necessary, the NTSB classified Safety Recommendations A-09-24 and -25 “Open—Acceptable

Response.”

1.18.2.2 Go-Around Callout

During the NTSB’s investigation of the February 18, 2007, accident involving a Shuttle

America (doing business as Delta Connection) Embraer ERJ-170, which overran the runway

during landing,

the NTSB concluded that both the PF and the PM should have the authority to

call for a go-around maneuver because one pilot might detect a potentially unsafe condition of

which the other pilot is unaware. As a result, the NTSB issued Safety Recommendation A-08-18,

which asked the FAA to do the following:

Require 14 Code of Federal Regulations Part 121, 135, and Part 91 subpart K

operators to have a written policy emphasizing that either pilot can make a

go-around callout and that the response to the callout is an immediate missed

approach.

On September 26, 2008, the FAA responded that it would publish a SAFO that discusses

the hazards associated with cockpit crewmembers who do not immediately and instinctively

react to a go-around callout while on an approach for landing. The FAA also stated that it would

revise FAA Order 8900.1 to include guidance to POIs to have all 14 CFR Part 121, 135, and 91

subpart K operators develop written policy that indicates that either pilot can make a go-around

callout and that the response to the callout is an immediate missed approach.

On June 22, 2009, the NTSB indicated that, pending the FAA’s revision of

FAA Order 8900.1 and its completion of a survey to ensure that all operators have developed and

implemented the recommended policy, Safety Recommendation A-08-18 was classified

“Open—Acceptable Response.” On March 1, 2010, the FAA issued SAFO 10005, “Go-Around

Callout and Immediate Response,” which provided information to operators and recommended

that they provide written policy to flight crews regarding go-around callouts. The other requested

action items have not yet been completed.

1.18.2.3 Low-Airspeed Alerting Systems

As a result of the February 12, 2009, Colgan Air accident,

the NTSB issued Safety

Recommendation A-10-12, which superseded Safety Recommendations A-03-53 and -54

and

asked the FAA to do the following:

See Runway Overrun During Landing, Shuttle America, Inc., Doing Business as Delta Connection Flight

6448, Embraer ERJ-170, N862RW, Cleveland, Ohio, February 18, 2007, Aircraft Accident Report

NTSB/AAR-08/01 (Washington, DC: National Transportation Safety Board, 2008).

<http://www.ntsb.gov/publictn/2008/AAR0801.pdf>.

NTSB/AAR-10/01.

Safety Recommendations A-03-53 and -54, which were issued on December 2, 2003, were classified

“Closed—Unacceptable Action/Superseded” on February 23, 2010, because of a lack of progress. Safety

Recommendation A-03-53 asked the FAA to do the following: “Convene a panel of aircraft design, aviation

operations, and aviation human factors specialists, including representatives from the National Aeronautics and

NTSB Aircraft Accident Report

For all airplanes engaged in commercial operations under 14 Code of Federal

Regulations Parts 121, 135, and [Part 91 subpart] K, require the installation of

low-airspeed alert systems that provide pilots with redundant aural and visual

warnings of an impending hazardous low-speed condition.

On June 22, 2010, the FAA responded that it planned to task the Aviation Rulemaking

Advisory Committee (ARAC) with developing recommendations for potential requirements to

address low-airspeed alerting systems. On January 25, 2011, the NTSB acknowledged that

asking the ARAC to address the issue of low-airspeed alerting systems may represent the first

step in an acceptable response to this recommendation. However, due to the FAA’s lack of

progress in this area, the NTSB classified Safety Recommendation A-10-12 “Open—

Unacceptable Response,” pending substantive action by the ARAC to address this

recommendation.

1.18.2.4 First Officer Assertiveness

As a result of its investigation of the February 19, 1996, accident involving Continental

Airlines flight 1943, a Douglas DC-9-32, which landed wheels up,

the NTSB issued Safety

Recommendations A-97-05 and -06, which asked the FAA to do the following:

Require all principal operations inspectors of 14 CFR Part 121 carriers to ensure

that the carriers establish a policy and make it clear to their pilots that there will

be no negative repercussions for appropriate questioning in accordance with crew

resource management techniques of another pilot’s decision or action. (A-97-05)

Require all principal operations inspectors of 14 CFR Part 121 carriers to ensure

that crew resource management programs provide pilots with training in

recognizing the need for, and practice in presenting, clear and unambiguous

communications of flight-related concerns. (A-97-06)

On October 30, 1998, the FAA issued AC 120-51C (currently AC 120-51E), which

emphasized the importance of all levels of management to support a safety culture that promotes

communication by encouraging appropriate questioning and that there will be no negative

repercussions for appropriate questioning of one pilot’s decision or action by another. The NTSB

also requested that the FAA issue a flight standards information bulletin (FSIB) to provide

guidance to POIs to ensure that operators are complying with the information presented in

Space Administration, to determine whether a requirement for the installation of low-airspeed alert systems in

airplanes engaged in commercial operations under 14 Code of Federal Regulations Parts 121 and 135 would be

feasible, and submit a report of the panel’s findings.” Safety Recommendation A-03-54 asked the FAA to do the

following: “If the panel requested in Safety Recommendation A-03-53 determines that a requirement for the

installation of low-airspeed alert systems in airplanes engaged in commercial operations under 14 Code of Federal

Regulations Parts 121 and 135 is feasible, establish requirements for low-airspeed alert systems, based on the

findings of this panel.”

In this accident, the captain rejected, without any discussion, the first officer’s go-around request, and the

first officer failed to challenge his decision. See Wheels-Up Landing, Continental Airlines Flight 1943, Douglas

DC-9-32, N10556, Houston, Texas, February 19, 1996, Aircraft Accident Report NTSB/AAR-97/01 (Washington,

DC: National Transportation Safety Board, 1997). <http://www.ntsb.gov/publictn/1997/AAR9701.pdf>.

NTSB Aircraft Accident Report

AC 120-51C. Based on the lack of action by the FAA to issue the FSIB, on February 23, 2000,

the NTSB classified these recommendations “Closed—Unacceptable Action.”

1.18.3 Airplane Equipment Options

1.18.3.1 Aircraft Performance Monitoring System

Since November 2005, new production ATR 42 and ATR 72 airplanes have been

equipped with an aircraft performance monitoring (APM) system

designed to assist flight

crews with monitoring airplane performance during operations in icing conditions. According to

ATR’s Cold Weather Operations publication,

ATR developed the APM to enhance a flight

crew’s ability to detect severe icing conditions after finding that flight crews do not always

adequately monitor the cues and follow the procedures established for responding to severe icing

encounters. The APM uses FDR-recorded parameters and other information

to calculate and

compare the airplane’s actual performance (with respect to airspeed and drag) with its expected

performance. The APM also computes the actual minimum icing and severe icing airspeeds. The

APM function is included in the airplane’s multipurpose computer and requires no additional

sensors or any calculation of atmospheric ice content.

If the APM detects that the minimum airspeed requirements are not being met or that

degraded airplane performance is present, the APM alerts the flight crew via aural chimes and

annunciator lights on each pilot’s panel. The system is designed to provide these alerts only

when the static air temperature is below 10° C, the airplane’s flaps and landing gear are retracted,

both engines are operating, and level 2 or 3 icing protection is engaged or an icing signal is

provided by the ice detector.

During cruise flight, if the APM detects that the airplane’s indicated airspeed is 10 kts

lower than the expected value, the blue “CRUISE SPEED LOW” annunciator lights will

illuminate to alert the flight crew to monitor the icing conditions and airspeed.

During climb,

cruise, or descent, if the APM detects that the airplane’s performance is becoming degraded, the

amber “DEGRADED PERF” annunciator lights will illuminate with a single aural chime to alert

the flight crew to apply the appropriate QRH procedures to verify that the deicing protection is

on, maintain red-bug airspeed plus 10 kts, and check for severe icing cues or abnormal handling.

The QRH states that, if such cues are absent, the flight may be continued and that icing

conditions and airspeed should be monitored. If the APM detects that the airspeed decays further

and the minimum icing airspeed is reached, the amber “INCREASE SPEED” annunciator lights

will flash with a single aural chime to alert the flight crew to increase the airspeed.

According to ATR, had the accident airplane been equipped with an APM, the flight crew

would have received the blue “CRUISE SPEED LOW” annunciator lights after about 10 minutes

The modification began with ATR 42-500 SN 641 and ATR 72-212A SN 726. The accident airplane, which

was older, was not equipped with an APM.

Cold Weather Operations: Be Prepared for Icing (2008).

The APM also uses aircraft weight information that must be input by the flight crew.

The light will extinguish once the appropriate airspeed is maintained.

NTSB Aircraft Accident Report

of cruise flight in the icing conditions at FL 180. During the descent through about 5,000 feet

msl, the flight crew would have received the blue “CRUISE SPEED LOW” annunciator lights

about 70 seconds before they selected the flap extension and the amber “DEGRADED PERF”

annunciator lights and single chime about 40 seconds before they selected the flap extension.

After the flight crew selected the flap extension, an APM would have provided no further

annunciator lights or chimes because the system requires a flaps-up configuration.

After the APM system was first introduced, on October 12, 2005, the Direction Générale

de l’Aviation Civile (DGAC) of France issued Recommendation Bulletin 2005/31(B) to “inform

the concerned authorities and operators of the new design improvement” and to “strongly

recommend” that ATR operators install the APM on their ATR fleets; the FAA did not issue a

similar advisory. In the bulletin, DGAC indicated that “no unsafe condition exists” for airplanes

that are not equipped with an APM provided that the flight crew applies the appropriate

procedures for detecting and responding to a severe icing encounter. However, DGAC noted that

an APM can help flight crewmembers recognize when they enter severe icing conditions and can

help them respond appropriately to such an encounter. On December 11, 2006, ATR issued All

Operators Message 42-72/2006/08 to notify all operators of the availability of the APM retrofits

and to recommend their installation. ATR also provided operators with information about the

APM in various publications and at several flight operations conferences worldwide.

On August 10, 2009, the European Aviation Safety Agency (EASA) issued

AD 2009-0170, effective August 24, 2009, that required all EASA-participating operators

(operators in the European Community, its Member States, and the European third countries) of

ATR 42 and 72 airplanes not originally equipped with the APM system to retrofit their airplanes

with the system within 72 months or by the second “C” check,

whichever occurs first. The

FAA took no similar action. As a result, APM system retrofits are optional for U.S. operators of

ATR 42 and ATR 72 airplanes not originally equipped with the system. According to ATR, of

the 780 ATR 42 and ATR 72 airplanes in service worldwide, 250 are equipped with APMs. Of

the 90 ATR 42 and ATR 72 airplanes in the United States, none are equipped with APMs.

1.18.3.2 Flap Malfunction or Asymmetry Light

Several ATR airplanes, including all ATR 42-200, -300, and -320 models (like the

accident airplane), were certificated without a cockpit light for a flap asymmetry or malfunction.

The certification basis that applies to these airplanes includes 14 CFR 25.699,

which states the

following:

there must be a means to indicate to the pilots the position of each lift or drag

device having a separate control in the cockpit to adjust its position. In addition,

an indication of unsymmetrical operation or other malfunction…must be provided

A “C” check is a heavy maintenance inspection performed at intervals based on manufacturer-specified

airplane usage parameters.

According to EASA, the original certification basis for the ATR 42-320 was Joint Aviation Requirements,

change 8 for paragraph 25.699, which is similar in content to 14 CFR 25.699, amendment level 25-23, effective

May 8, 1970.

NTSB Aircraft Accident Report

when such indication is necessary to enable pilots to prevent or counteract an

unsafe flight…condition.

Certification testing demonstrated that a flap asymmetry in these airplanes (which is

limited to less than 10° by the airplanes’ flap asymmetry detection system)

did not constitute

an unsafe flight condition because the airplanes remained controllable. Later-model ATR

airplanes, including all ATR 42-400 and -500 and all ATR 72-101, -102, -201, -211, -212, and

-212A models, have a cockpit light that indicates a flap asymmetry. The airplane manufacturer

stated that the asymmetry light was added as part of an upgrade to the airplanes’ flight

management system. The flap asymmetry light was designed to illuminate amber any time that

an asymmetry greater than 6.7° occurred. Nearly all transport-category airplanes, with the

exception of the ATR 42-200, -300, and -320 models and the Cessna 500, 550, 553, 560, and

560XL models, are certificated with some type of flap malfunction indicator in the cockpit, some

of which specifically indicate flap asymmetry.

1.18.3.3 Ice Evidence Probe

Since 1995, new-production ATR airplanes have been equipped with an IEP, which is

mounted so that it is visible from the left-side cockpit window and is considered the primary

means for flight crews to monitor ice accumulation.

The IEP provides an airfoil-shaped surface

upon which ice accumulation can be visually detected and is designed to retain ice on its surface

until the entire aircraft is free of ice, thus helping flight crews determine when critical surfaces of

the airplane may be free of ice. The IEP is automatically illuminated for use during night

operations when the airplane’s navigation lights are turned on. For all ATR airplanes that were

not originally equipped with IEPs, ATR provided the operators with the service bulletin and kits

necessary for installing IEPs free of charge. These kits were distributed in 1995, and the

installation was optional.

The accident airplane was not equipped with an IEP.

FAA personnel indicated that most airplanes’ flap asymmetry detection systems are similar in that they limit

the amount of asymmetry that can occur.

Cold Weather Operations: Be Prepared for Icing (2008).

NTSB Aircraft Accident Report

2. Analysis

2.1 General

The captain and the first officer were certificated in accordance with Federal regulations

and were current and qualified in accordance with Empire Airlines’ training requirements. The

investigation found no evidence that the pilots’ performance was affected by any behavioral or

medical condition, or by the use of alcohol or drugs.

The airplane was loaded within weight and CG limits and was maintained in accordance

with 14 CFR Part 121. Although Empire Airlines could not immediately locate the left engine’s

AD records, work orders and other records were located that indicated the AD compliance items

had been performed. The investigation found no evidence that indicated any mechanical anomaly

with the engines, and FDR data indicated that the engines’ performance was consistent with

normal operations.

The flight experienced two separate periods of airframe ice accretion, the first during

cruise flight at FL 180 and the second during descent below 6,000 feet msl. The flight crew

selected ice protection during both icing encounters, and no evidence indicated any mechanical

anomalies with the airplane’s deicing and anti-icing systems. According to the captain, following

the first icing encounter, during the descent above 6,000 feet msl, “substantial amounts of ice”

came off the airplane. The shedding of ice is consistent with the presence of a layer of air with

temperatures above freezing between 6,000 and 11,000 feet msl due to a temperature inversion.

Further, performance calculations based on FDR data indicated that, during the descent between

14,000 to about 5,000 feet msl, the airplane’s performance was as expected for an ice-free

airframe. During the approach into LBB, the airplane began to accumulate ice again when it

descended into moderate icing conditions resulting from light freezing drizzle, a weather

condition that was outside its icing certification envelope as specified in 14 CFR Part 25,

Appendix C.

During the approach as the airplane descended through about 4,800 feet msl, the FDR

recorded that a flap asymmetry occurred about 1 second after the cockpit flap position lever was

positioned to command 15° flaps. The FDR recorded no operational anomalies with the

airplane’s hydraulic systems, and the flight crew reported none. Performance data indicated that,

when the flight crew commanded 15° flaps, a flap asymmetry occurred with the left flaps

extending 8° to 10° and the right flaps not extending. The data further indicated that the flaps

returned to a symmetric state (about 4.5°) about 25 seconds before ground impact.

Based on the circumstances of the accident flight, a number of scenarios could have

resulted in a flap asymmetry, and the investigation evaluated the possibility that a mechanical

anomaly, flap actuator jamming, mechanical restriction of the movement of the right flaps, or

hydraulic fluid contamination may have occurred. Postaccident impact and fire damage to the

right flaps limited the extent to which preimpact system integrity could be determined

For information about the NTSB’s longstanding concerns about the operation of airplanes in weather

conditions that are outside their icing certification envelope, see section 2.4.1.

NTSB Aircraft Accident Report

(components from the left flaps sustained less damage during the accident sequence and the left

flaps’ actuators functioned when tested). X-ray examinations and disassembly of all four flap

actuators (for both the left and right flaps) showed that their internal components were intact at

the time that the asymmetry occurred, that hydraulic fluid (or evidence of hydraulic fluid) was

present in each actuator, and that no evidence existed of any mechanical anomaly or jamming of

the actuators. Examination of the available components from the right flaps revealed no evidence

that any mechanical restriction of the movement of the flaps had occurred with respect to those

components (thermal damage precluded examination of the entire system).

According to ATR, ice formation on the flap surfaces could not result in a delayed

extension of the flaps because in all types of icing conditions, runback icing does not exceed 15

to 20 percent of the wing chord on the upper wing surface and 30 to 40 percent of the wing chord

on the lower wing surface and that these limits are upstream of the flaps. Further, ATR noted that

there were no reports of a flap asymmetry due to ice formation in the operational history of the

aircraft make/model. However, because the airplane was operating outside of its icing

certification envelope at the time of the accident and runback ice was found on the lower surface

of a separated section of the right aileron, the investigation could not rule out the possibility that

ice could have formed and either delayed or restricted the flap extension and resulted in a flap

asymmetry.

Testing of recovered hydraulic fluid found that solid particles, which testing found to be

consistent with heated hydraulic fluid, were found in the hydraulic fluid lines and the connector

port for the right outboard flap actuator. No similar particles were found in any fluid samples

from the airplane’s hydraulic reservoir or from the examined hydraulic system filters. Therefore,

the identified solid particles appeared to have changed state only after exposure to the high

temperatures of the postimpact fire.

The investigation considered that the right inboard flap actuator (installed as original

equipment on the airplane) may have sustained interior damage during the airplane’s 1998 right

engine fire event that may have gone undetected and resulted in a malfunction of the actuator.

The investigation also considered that hydraulic fluid contamination (such as from foreign

particles or air) could result in a flap actuation delay. No evidence of any preexisting condition

was discovered. However, because of the extent of the impact and postcrash fire damage to the

right flaps’ actuators and the hydraulic system components, the investigation could not determine

if any evidence of either condition may have been present before the accident. The NTSB

concludes that, because of the extent of the impact and postaccident fire damage to the right flaps

and flap actuators, the reason for the airplane’s flap asymmetry could not be determined.

The following analysis describes the accident sequence and examines the safety issues

associated with the flight crew’s performance during the approach in response to the flap

anomaly, the dispatch of the flight into freezing drizzle conditions, the communication of flight

crew and HAZMAT information to the emergency response personnel, airplane equipment

considerations, and simulator-based training for pilots who fly in icing conditions.

NTSB Aircraft Accident Report

2.2 Accident Sequence

During the accident approach, at 0434:24, the first officer was flying using the autopilot

and called for a flap setting of 15°, which is the appropriate approach flap setting for the

ATR 42; however, a flap asymmetry occurred. During a postaccident interview, the first officer

stated that she noticed that the airplane did not decelerate as expected (and briefly accelerated);

however, she did not say anything to the captain about her observation. The airplane’s failure to

decelerate is consistent with a flap asymmetry because the abnormal condition provided less lift

and drag than the desired 15° approach flap configuration. Further, the presence of airframe ice

accretion also affected the drag forces on the airplane. As a result, the engine power settings that

would have been appropriate during a typical approach (with 15° flaps and no ice contamination)

would not have applied to the accident approach. However, the FDR showed that at 0434:54,

about 30 seconds after the asymmetry occurred, the first officer reduced engine power to about

3 percent torque, which was likely an effort to reduce the airspeed; as the airplane continued its

descent on autopilot, only a minimal increase in engine power was commanded.

About 40 seconds after the asymmetry occurred, the captain observed a problem with the

flaps; at this time, the airplane was just outside the LOM/FAF at an altitude of about 1,400 feet

agl. In response, the captain began a nonstandard procedure to troubleshoot the flap problem and

did not verbalize his plan of action with the first officer, who continued flying the approach. The

airplane’s airspeed decreased throughout the approach, resulting in an increasing AOA, and the

stick shaker activated when the airplane was about 900 feet agl at an airspeed of 125 kts. The

airplane performance study found that the stick shaker activated at the setting prescribed for

flight in icing conditions.

The stick shaker disconnected the autopilot at 0435:29, and the first officer began to

manually fly the approach, increasing engine power to about 70 percent torque. Control force

estimates showed that the first officer was applying about 35 to 40 lbs of rudder pedal pressure

and about 13 lbs of control wheel force to manually counter the flap asymmetry. However, under

manual control, the airplane deviated high and to the right of the runway 17R ILS glidepath,

indicating that the first officer may not have immediately recognized the extent of the airplane’s

right banking tendency induced by the flap asymmetry. As the deviation increased, at 0435:40,

the first officer asked the captain, “should I go around?” The captain replied, “no” and instructed

the first officer to “keep descending.” About 4 seconds later, when with a strained voice the first

officer stated, “we’re getting close here,” the captain asked the first officer if she wanted him to

take the controls, and she replied that she did.

After the accident, the captain stated that he had not heard the autopilot disconnect alert

and was surprised to see that the first officer was manually flying the airplane. No autopilot

disconnect alert was heard on the CVR. Postaccident testing of the autopilot disconnect alerting

system revealed no malfunctions. On the ATR 42-320, the aural stall warning has priority over

the autopilot disconnect aural alert. By the time that the autopilot disconnect aural alert would

have sounded (about 1.1 seconds after the aural stall warning), the first officer likely responded

For more information about the captain’s actions and the flight crew’s training and procedures for responding

to flap anomalies and the stick shaker, see section 2.3.1.

NTSB Aircraft Accident Report

to the stall warning by pushing the autopilot disconnect push button, which silenced the pending

autopilot disconnect alert.

When the captain took control, the airplane was about 700 feet agl at 143 KIAS, high and

to the right of the approach course. At 0436:00, about 10 seconds after the captain took control,

the forces exerted on the cockpit controls peaked with maximum forces on the control column,

control wheel, and rudder pedals of 36 lbs, 29 lbs, and 112 lbs, respectively.

When these peak

forces were applied, the airplane was descending through about 500 feet agl (the reported

ceiling) and likely emerged from the clouds, allowing the captain to establish visual contact with

the runway. At this time, the airplane was more than 300 feet right of the localizer course, and

the captain’s control inputs were consistent with a course correction to align the airplane with the

runway. About 2 seconds later, the flaps returned to a symmetric state, and control input to

counteract the flap asymmetry was no longer required.

Immediately after the captain exerted the peak forces on the cockpit controls, about

27 seconds before impact, the aural stall warning and the stick shaker activated for the second

time, concurrent with the TAWS warning, “pull up. pull up.” Procedures for responding to either

the stick shaker or the TAWS warning require the immediate application of maximum engine

power; however, 2 seconds later, the captain reduced the airplane’s engine power to about

10 percent torque, and he did not call for maximum engine power until 15 seconds later at

0436:17. As the captain continued the approach, the airspeed again decreased below the

minimum safe approach speed.

At 0436:18, about 9 seconds before impact, the pilot began to pull back on the control

column for several seconds. In response to the pilot’s pull on the control column, the pitch

attitude, AOA, and load factor all began increasing. At 0436:19, the aural stall warning and the

stick shaker activated for the third time. The airspeed when the stall warning occurred was about

124 kts, approximately 19 kts below the minimum safe airspeed for the approach of 143 kts.

About 2 seconds after the stall warning occurred, the airplane experienced an uncommanded roll

followed by a series of roll, yaw, and pitch oscillations, which continued until ground impact.

No evidence existed of airplane performance behavior consistent with an aerodynamic

stall or with a loss of lateral control until after 0436:20. Further, no evidence existed of a sudden

change in aileron hinge movement throughout the approach. The airplane performance study and

simulations determined that the performance degradation due to ice accretion never exceeded the

airplane’s thrust performance, nor would it have exceeded the airplane’s flight control

capabilities if the minimum safe airspeed had been maintained. Therefore, the NTSB concludes

that the airplane was controllable with the flap asymmetry and airframe ice contamination and

could have been maneuvered and landed safely if the minimum safe airspeed had been

maintained.

If the captain had responded appropriately to the concurrent stall and TAWS warnings

that occurred 27 seconds before impact by immediately initiating a go-around maneuver when

the airplane was at an altitude of about 500 feet agl, he likely would have been able to arrest the

According to 14 CFR 25.143(d), the maximum control forces allowable for short-term application of pitch

and roll, when two hands are available, and yaw are 75 lbs, 50 lbs, and 150 lbs, respectively.

NTSB Aircraft Accident Report

airplane’s descent, increase its airspeed, and avoid an aerodynamic stall. The NTSB concludes

that the captain’s failure to immediately respond to the aural stall warning, the stick shaker, and

the TAWS warning resulted in his inability to arrest the airplane’s descent and avoid impact with

the ground.

2.3 Human Performance

Under normal conditions, the approach-to-landing is a high-workload phase of flight for

pilots. Required tasks include controlling, maneuvering, and configuring the airplane;

performing final landing checks; and evaluating the airplane’s performance relative to the

landing criteria. Pilots are expected to capture and prioritize critical information, analyze that

information, and take appropriate action. Performing these tasks can become increasingly

difficult as workload increases beyond routinely experienced levels, and, during the accident

approach, both the icing conditions and the flap anomaly combined to increase the flight crew’s

workload.

Research demonstrates that emergency situations increase workload and require

additional effort to manage effectively because of the stress involved, human cognitive

limitations, and the lack of opportunity for pilots to practice these skills on a regular basis

compared to the skills used in normal operations.

The NASA emergency and abnormal

situations study,

which analyzed the NTSB’s findings from its 1994 safety study that reviewed

flight crew-involved, major accidents,

also noted that, during high-workload conditions,

performance deficiencies, including narrowing of attention and impairment of short-term

memory, could result from inherent limitations in cognitive processes and the effect of stress on

human performance.

Training, experience, and SOPs are developed to mitigate errors, especially during

dynamic and high-workload situations. However, after the flight crewmembers recognized that a

flap problem existed, they failed to follow their training and did not perform a go-around

maneuver and reference the QRH checklist for the flap anomaly procedure; they did not comply

with Empire Airlines’ procedures (outlined in the GOM) that address abnormal situations. As a

result, a compounding situation unfolded in which the flight crew missed several opportunities to

apply the correct SOPs to ensure the safety of the flight.

SOPs are universally recognized as basic to safe aviation operations. Well-designed cockpit

procedures are an effective countermeasure against operational errors, and disciplined compliance

with SOPs, including strict checklist discipline, provides the basis for effective crew coordination

and performance. Operational data confirm the importance of strict compliance with SOPs for safe

operations.

For example, industry data show that pilots who intentionally deviated from SOPs were

R.K. Dismukes, G.E. Young, and R.L. Sumwalt, “Cockpit Interruptions and Distractions: Effective

Management Requires a Careful Balancing Act,” ASRS Directline, vol. 10 (1998), pp 4-9.

B.K. Burian, I. Barshi, and K. Dismukes, The Challenge of Aviation Emergency and Abnormal Situations,

NASA Technical Memorandum 2005-213462 (Moffett Field, California: National Aeronautics and Space

Administration, 2005).

See A Review of Flight Crew-Involved, Major Accidents of U.S. Air Carriers, 1978 through 1990, Safety

Study NTSB/SS-94/01 (Washington, DC: National Transportation Safety Board, 1994).

NTSB Aircraft Accident Report

three times more likely to commit other types of errors, mismanage errors, and find themselves in

undesired situations compared with pilots who did not intentionally deviate from procedures.

According to AC 120-71A, in its study of controlled flight into terrain accidents, the Commercial

Aviation Safety Team, which included FAA, industry, and union representatives, found that almost

50 percent of the studied accidents related to the flight crew’s failure to adhere to SOPs or the

certificate holder’s failure to establish adequate SOPs. The NTSB has repeatedly cited casual cockpit

discipline and inadequate compliance with SOPs as contributing factors to accidents.

2.3.1 Captain’s Nonstandard Responses

After the captain stated that the airplane had “no flaps,” no further discussion occurred as

to what actions should be taken, and no reference was made to the QRH. Instead, the captain

retrieved and used a flashlight to check the circuit breakers located behind the first officer’s seat

while the first officer continued to fly the approach. During postaccident interviews, the captain

indicated that he checked the circuit breakers because he felt that he needed to determine what

the malfunction was before he could respond with the appropriate checklist. The first officer

stated that she was a little concerned about the captain’s actions and that she thought that he

should have been more focused on the tasks associated with landing the airplane. However, at no

time during the flight did she verbalize her concerns.

Both the captain and the first officer had been trained to perform a go-around maneuver

and reference the QRH if a flap problem occurred during an approach. Based on the flight

crew’s training and the low altitude at which they first noticed the flap anomaly, the appropriate

response would have been to maintain control of the airplane, perform a go-around manuever

(climb the airplane to a safe altitude according to the missed approach procedure), and refer to

the QRH procedure for a flap anomaly. If a go-around maneuver had been performed, after

climbing to 6,000 feet msl, the airplane would have exited icing conditions and reentered a layer

of warmer air with temperatures above freezing. The flight crew could then have completed the

QRH procedure and determined whether to attempt another approach into LBB or proceed to the

flight’s alternate airport, which had no reported precipitation.

Although the captain’s observation that the airplane had “no flaps” was inaccurate

because the airplane actually had a flap asymmetry, the inaccuracy was irrelevant because the

same QRH procedure applied for both a no-flaps condition and a flap asymmetry. The pilots

were trained to verify the flap conditions by looking in the cockpit at the flap position indicator

and out the side windows at the flap position fairings. The procedure would then have been to

The data came from the LOSA Collaborative, which is a network of researchers, safety professionals, pilots,

and airline representatives collaborating to provide, among other things, oversight and implementation of line

operational safety audits and a forum of information exchange regarding these audits. More information is available

online at <http://losacollaborative.org/>.

For more information, see (a) Collision with Trees and Crash Short of the Runway, Corporate Airlines

Flight 5966, BAE Systems BAE-J3201, N875JX, Kirksville, Missouri, October 19, 2004, Aircraft Accident Report

NTSB/AAR-06/01 (Washington, DC: National Transportation Safety Board, 2006); (b) Attempted Takeoff From

Wrong Runway, Comair Flight 5191, Bombardier CL-600-2B19, N431CA, Lexington, Kentucky, August 27, 2006,

Aircraft Accident Report NTSB/AAR-07/05 (Washington, DC: National Transportation Safety Board, 2007); and

October 14, 2004, Aircraft Accident Report NTSB/AAR-07/01 (Washington, DC: National Transportation Safety

Board, 2007).

NTSB Aircraft Accident Report

place the cockpit flap control lever in the position nearest to the actual position of the flaps and

next perform a reduced-flaps landing procedure (in this case, a no-flaps landing). The no-flaps

landing procedure included briefing the no-flaps approach procedure and resetting the airspeed

bugs.

The NTSB acknowledges that recognizing the cockpit warning indicators, identifying the

nature of any problems, and choosing a response procedure requires considerable attention. As

noted beneath the QRH Table of Contents, before a flight crew performs a procedure, the flight

crew “must assess the situation as a whole, taking into consideration the failures, when fully

identified, and the flight constraints imposed.” According to the GOM, in the event of an

inoperative or malfunctioning flight control system, if the airplane is controllable, the flight crew

“should complete only the applicable checklist procedures” and “not attempt any corrective

action beyond those specified.”

Although the NTSB recognizes that the PIC has authority to deviate from prescribed

procedures in the interest of safety, no indication existed that the captain considered this to be an

emergency condition, and the airplane was controllable. Further, the captain identified the flap

anomaly to the extent necessary to select the applicable checklist procedure. The NTSB

concludes that the captain was adequately trained on how to respond to flap anomalies and that

the captain’s statement that the airplane had “no flaps” indicates that he had sufficient

information to recognize that he should immediately perform a go-around maneuver and apply

the appropriate procedure from the QRH that applied to all flap problems.

Both the captain and the first officer had been trained to perform a go-around maneuver

during an approach in which a stick shaker or TAWS warning occurred or an approach that

became unstabilized. Multiple cues existed that indicated the approach was unstabilized,

including excessive deviation from the glidepath (more than 300 feet right at 500 feet agl), sink

rate greater than 1,000 fpm, and airspeed less than the required approach speed. The captain

indicated that he made a judgment call to continue the landing because he was concerned about

the flap anomaly and the icing conditions. However, the captain had extensive experience flying

in icing conditions and was capable of assessing the effects of ice accretion on the airplane’s

performance. Because the airplane had shed ice from its first icing encounter during the descent

into LBB, the captain should have recognized that a climb to 6,000 feet msl would have allowed

the flight to exit icing conditions. Further, the flight crew had selected level 3 ice protection for

the approach, and neither flight crewmember indicated that the ice accretion itself was

worrisome before the flap anomaly occurred. The NTSB concludes that although the captain

indicated that he was concerned about the icing conditions, the presence of airframe ice accretion

within the capabilities of the airplane’s systems does not negate the importance of adhering to

SOPs and performing a go-around maneuver to respond to the multiple cues associated with an

unstabilized approach including excessive deviation from the glidepath, sink rate greater than

1,000 fpm, and airspeed less than the required approach speed. Further, the NTSB concludes

that, had the captain complied with SOPs in response to the flap anomaly, unstabilized approach,

stick shaker, and TAWS warning and initiated a go-around maneuver, the accident likely would

not have occurred.

NTSB Aircraft Accident Report

2.3.2 Flight Crew’s Failure to Monitor Airspeed

The dynamic nature of the approach phase of flight requires flight crews to monitor

airspeed, and the presence of icing conditions requires an increased awareness of the airspeed

because of the detrimental effects that icing can have on aircraft performance. According to the

ATR 42 pilot handbook, during an ILS approach, the PM is to call out any deviations from

normal altitude, airspeed, or descent rates. The QRH states that, during a non-normal event, the

PF is responsible for the engine power levers, flightpath and airspeed control, aircraft

configuration, and navigation, and the PM is responsible for completing the appropriate checklist

items, actions on the overhead panel, condition levers, and communications.

During the 26 seconds from when the captain identified a flap problem to the first

stick-shaker event, the airplane’s airspeed decreased from about 155 kts (above the red-bug

airspeed of 143 kts for a 0° flap approach in icing conditions) to 125 kts, and the first officer

applied only a minimal adjustment of engine power during this time. As the airspeed decayed,

neither flight crewmember made any verbal reference about airspeed (actual or target), and at no

time did the captain call out the localizer, glideslope, airspeed, or vertical speed deviations.

Each pilot had two indicators available to assist with airspeed monitoring. First, each

pilot had a fixed-scale, moving-pointer analog airspeed indicator. Each airspeed indicator had

four airspeed bugs—yellow, white, red, and internal—that were manually set by each pilot based

on predetermined airspeeds for the flight conditions.

100

The airspeed bugs for approach and

landing in icing conditions should have been set at 121, 123, 143, and 116 kts, respectively. The

captain erroneously briefed the bugs to be set at takeoff speeds in icing conditions, specifically

112, 123, 143, and 106 kts. The captain’s bugs were set at 110, 124, 145, and 109 kts, and the

first officer’s bugs were set at 112, 124, 144, and 106 kts, respectively. Second, a fast/slow

indicator was located on the left side of each pilot’s EADI. The fast/slow indicator alerts the pilot

of the airplane’s actual airspeed in relation to the target airspeed set by the internal bug.

When the pilots realized that a flap anomaly had occurred, the appropriate airspeed to

maintain would have been the red bug, or 143 kts. The crew could have reset the internal bug to

this airspeed to have the additional airspeed cue from the EADI fast/slow indicator, but they did

not do so. Previous accidents

101

have shown that pilots can become distracted from flying duties

when an emergency or abnormal situation occurs, and literature suggests that “one of the biggest

hazards of ‘abnormals’ is becoming distracted from other cockpit duties.”

102

While flying the

100

For more information about the airspeed bugs, see sections 1.6.4.1 and 1.17.1.4.

101

(a) On December 29, 1972, a Lockheed L-1011 crashed when the flight crewmembers failed to monitor their

flight instruments after becoming distracted by a malfunctioning landing gear position indicator, resulting in

101 fatalities. See Eastern Air Lines, Inc., L-1011, N310EA, Miami, Florida, December 29, 1972, Aircraft Accident

Report NTSB/AAR-73/14 (Washington, DC: National Transportation Safety Board, 1973). (b) On June 9, 1995, a

de Havilland DHC-8 crashed when the captain failed to monitor the approach after becoming distracted by first

officer’s attempts to reconcile a landing gear malfunction, resulting in 4 fatalities and 14 serious injuries. See de

Havilland DHC-8, ZK-NEY, Controlled Flight into Terrain, near Palmerston North, 9 June 1995, Investigation

Report 95-011 (Wellington, New Zealand: Transport Accident Investigation Commission, 1997).

<http://www.taic.org.nz> (accessed March 21, 2011). (c) On April 19, 2008, a Cessna 510 overshot the runway

during landing in Carlsbad, California, after the pilot became distracted by a flickering primary flight display and

misjudged the airspeed and landing distance. The report for this accident, NTSB case number LAX08FA117, is

available online at <http://www.ntsb.gov/aviationquery/index.aspx>.

102

Dismukes, Young, and Sumwalt, pp 4-9.

NTSB Aircraft Accident Report

approach, the first officer was likely distracted from monitoring the instruments by the flap

anomaly, the captain’s nonstandard actions involving the circuit breakers, and the control force

inputs needed to maintain control of the airplane because of the flap asymmetry. Further, for the

captain to check circuit breakers behind the first officer’s seat, he would need to turn away from

the instrument panel, a position from which monitoring the instruments was not possible. The

NTSB concludes that the first officer’s failure to maintain airspeed while acting as the PF likely

resulted from being distracted by the flap anomaly, the captain’s actions in response to it, and the

control force inputs needed to maintain aircraft control. Further, the NTSB concludes that the

captain’s failure to call out the first officer’s airspeed deviations resulted directly from his

preoccupation with performing an inappropriate, nonstandard procedure in response to the flap

anomaly.

2.3.2.1 Airspeed References

As previously mentioned, the airspeed bugs were not set properly because the captain did

not correctly brief the airspeeds for an approach in icing conditions; instead, he erroneously

briefed the airspeeds for a takeoff in icing conditions. It is not clear why neither crewmember

recognized the captain’s error; however, at the time of the accident, Empire Airlines did not have

a procedure in place for flight crews to crosscheck and verify the setting of airspeed bugs.

Empire Airlines has since implemented a crosscheck procedure. Also, because the captain did

not perform the QRH procedure after identifying the flap anomaly, he did not brief the new

approach speeds, and the flight crew did not reset the airspeed bugs accordingly.

Because the flight crew failed to properly set the internal airspeed bug, the EADI

fast/slow indicator, which indicates the difference between the internal bug airspeed and the

airplane’s actual airspeed, provided indications that did not apply to the accident approach.

However, the captain had correctly stated during his initial briefing that the airspeed for the red

bug, which indicates the approach airspeed for 0° flaps in icing conditions, was 143 kts, and both

red bugs were set near that value (144 and 145 kts). Therefore, the NTSB concludes that,

although some of the airspeed bugs (including the internal bug) were not set to the appropriate

approach airspeeds and were not reset following recognition of the flap anomaly, the flight crew

had a sufficient reference to maintain the minimum safe airspeed because the airspeed for a

no-flap approach in icing conditions was correctly briefed as 143 kts, and the red airspeed bugs

were set near that value.

The airspeed indicator is a visual cue that can convey meaning to flight crewmembers

only when they look at the indicator. As discussed previously, the first officer was distracted

from monitoring the visual airspeed indications, and the captain’s attention was focused away

from the instrument panel. Evidence from this accident and others indicates that, in

high-workload situations, a visual-only cue may not capture a pilot’s attention due to human

sensory limitations.

The airplane’s stall protection system, which included an aural warning and tactile stick

shaker, is designed to comply with Federal airworthiness standards, specifically

NTSB Aircraft Accident Report

14 CFR 25.207,

103

that require airplanes be equipped to provide flight crews with a clear and

distinctive stall warning at an airspeed that is about 7 percent higher than stall airspeed. When

the flight crew became distracted by the flap anomaly and failed to detect that the airplane was

approaching a stall, the stall protection system activated as designed and warned the flight crew

of the decaying airspeed. However, the airspeed began to decay unnoticed much earlier during

the approach. The airspeed indicator should have provided the pilots with adequate time to detect

the ensuing low-airspeed regime and respond appropriately. However, distraction and workload

considerations may have made it difficult for the flight crew to visually detect these cues. The

NTSB concludes that reliance upon flight crew vigilance and stall warning systems may be

inadequate to prevent hazardous low-airspeed situations and that, had a low-airspeed alerting

system been installed on the airplane, it may have directed the flight crew’s attention to the

decaying airspeed earlier and provided an opportunity to take corrective action before the stall

protection system activated.

The circumstances of this accident and the history of accidents involving flight crews’

lack of low-airspeed awareness, most recently the flight crew of Colgan Air flight 3407,

104

suggest that flight crew vigilance and existing stall warnings may be inadequate to reliably

prevent hazardous low-airspeed situations. In the Colgan Air accident, the pilots likely did not

see the rising low-airspeed cue on the indicated airspeed display, the downward-pointing trend

vector, or the airspeed indications change to red. As a result, the NTSB concluded that an aural

warning in advance of the stick shaker would have provided a redundant cue of the visual

indication of the rising low-airspeed cue and might have elicited a timely response from the

pilots before the onset of the stick shaker. Therefore, the NTSB issued Safety Recommendation

A-10-12,

105

which recommended that, for all airplanes engaged in commercial operations under

14 CFR Parts 121, 135, and 91 subpart K, the FAA require the installation of low-airspeed

alerting systems that provide pilots with redundant aural and visual warnings of an impending

hazardous low-airspeed condition. Because of the FAA’s inactivity regarding previous

recommendations to require the installation of low-airspeed alerting systems, the NTSB

classified Safety Recommendation A-10-12 “Open—Unacceptable Response” when it was

issued on February 23, 2010.

Although the FAA has tasked the ARAC with developing recommendations for potential

requirements to address low-airspeed alerting systems, progress remains slow, and the

recommendation is still classified “Open—Unacceptable Response.” This accident once again

103

At the time the ATR 42 was certificated in 1988, 14 CFR 25.207(c) stated that “the stall warning must begin

at a speed exceeding the stalling speed (i.e., the speed at which the airplane stalls or the minimum speed

demonstrated, whichever is applicable under the provisions of 14 CFR 25.201(d)) by seven percent or at any lesser

margin if the stall warning has enough clarity, duration, distinctiveness, or similar properties.” The current version

of 14 CFR 25.207(c) states that “when the speed is reduced at rates not exceeding one knot per second, stall warning

must begin, in each normal configuration, at a speed, V

, exceeding the speed at which the stall is identified in

accordance with 14 CFR 25.201(d) by not less than five knots or five percent CAS, whichever is greater. Once

initiated, stall warning must continue until the angle of attack is reduced to approximately that at which stall warning

began.”

104

NTSB/AAR-10-01.

105

For more information about Safety Recommendation A-10-12, see section 1.18.2.3.

NTSB Aircraft Accident Report

demonstrates the need for low-airspeed alerting systems and provides additional support for this

recommendation.

106

2.3.2.2 Flight Crew Monitoring Skills and Techniques

The accident flight crew’s deficiencies during the approach demonstrate the importance

of active monitoring skills. However, the flight crew’s CRM training did not specifically address

the development of active monitoring skills.

107

The NTSB has long recognized the importance of

flight crew monitoring skills in accident prevention. In 1994, the NTSB published a safety

study

108

that reviewed 37 flight crew-related, major accidents involving U.S. air carriers and

found that, for 31 of these accidents, inadequate monitoring and/or crosschecking had occurred.

The safety study found that flight crewmembers frequently failed to recognize and effectively

draw attention to critical cues that led to the accident sequence.

As a result of this safety study, the NTSB issued safety recommendations that resulted in

upgrades of FAA training guidance, including revisions to AC 120-51B (currently AC 120-51E),

“Crew Resource Management Training,” and AC 120-71 (currently AC 120-71A), “Standard

Operating Procedures for Flight Deck Crewmembers,” to include references to the importance of

flight crew monitoring functions. Subsequently, the NTSB issued Safety Recommendation

A-07-13,

109

which asked the FAA to require that all pilot training programs contain modules that

teach and emphasize monitoring skills and workload management and include opportunities to

practice and demonstrate proficiency in these areas.

On May 17, 2007, the FAA stated that it would consider identifying in its work program

a list of required inspections, reemphasizing to the regional and FSDO managers the need to

validate the training and to verify its effectiveness. However, the FAA’s lack of progress in this

area and the February 12, 2009, Colgan Air accident

110

in which the flight crew failed to monitor

airspeed prompted the NTSB to reiterate Safety Recommendation A-07-13 and reclassify it

“Open—Unacceptable Response” on February 23, 2010. On the same date, the NTSB issued

Safety Recommendation A-10-10,

111

which asked the FAA to require 14 CFR Part 121, 135, and

91 subpart K operators to review their SOPs to verify that they are consistent with the flight crew

monitoring techniques, as described in AC 120-71A. On June 22, 2010, the FAA responded that

it was considering issuing a SAFO or an InFO bulletin to address this recommendation. On

January 25, 2011, the NTSB indicated that, although such action may represent part of an

acceptable response to this recommendation, the FAA must also ensure that the guidance

provided in documents like SAFOs is implemented. This accident again demonstrates the

importance of flight crew monitoring skills in accident prevention and provides additional

support for Safety Recommendations A-07-13 and A-10-10.

106

For information about ATR’s APM technology that provides icing-specific airspeed alerts, see section 2.6.1.

107

For information about other CRM topics, see section 2.3.3.

108

NTSB/SS-94/01.

109

For more information about Safety Recommendation A-07-13, see section 1.18.2.1.

110

NTSB/AAR-10/01.

111

For more information about Safety Recommendation A-10-10, see section 1.18.2.1.

NTSB Aircraft Accident Report

2.3.3 Crew Resource Management

During the accident flight’s initial descent, the flight crew demonstrated several instances

of good CRM; the captain completed the descent checklist and immediately began the approach

checklist in anticipation of the first officer’s needs. However, after the flap anomaly was

recognized, the flight crew’s CRM deteriorated to the extent that the captain did not provide

adequate leadership in response to the anomaly. Further, the first officer was not assertive in

voicing her concerns about the captain’s nonstandard response or his decision to continue the

unstabilized approach. The captain is responsible for setting the tone in the cockpit for the entire

flight, and this responsibility was even more critical because the flight crew was faced with an

abnormal situation.

The accident captain had about 13,935 total flight hours (2,052 hours of which were in

the ATR 42) and extensive experience flying in icing conditions. He had previously served as a

check airman on the company’s single-pilot Cessna 208 airplanes. The first officer had about

2,109 total flight hours (130 hours of which were as SIC in the ATR 42) and very limited

experience flying in icing conditions. The captain had been with Empire Airlines for more than

20 years.Thus, he should have been very familiar with the company’s SOPs, and the accident

flight crew should have been able to benefit from the captain’s extensive experience with the

company, in ATR 42 airplanes, and with flying in icing conditions. However, the captain failed

to command control of the situation and follow SOPs. Following identification of the flap

anomaly, the captain did not communicate to the first officer his intentions for managing the

situation. Instead, the captain pursued his own plan of action without providing any guidance to

the first officer or performing the duties required of him as the PM, which could have helped the

first officer manage the airplane during the approach.

The large disparity between the captain’s and the first officer’s flying experience, their

experience in the ATR 42, and their experience in icing conditions likely created a steep

authority gradient in the cockpit. The flight crewmembers who flew with the captain stated in

interviews that, overall, he was very experienced, a good pilot, and a good captain. He was

referred to as a “guru.” The issue of cockpit authority gradient has been examined since the late

1970s and has been evident in a number of major aviation accidents, including the March 27,

1977, collision of two Boeing 747s at Tenerife Airport, Los Rodeos, Spain.

112

Evidence has

suggested that too steep of a gradient can reduce flight crew performance, hinder

communication, and increase errors. For example, a review of U.S. Navy mishaps between 1980

and 1990 revealed that flight crews with an authority gradient of one rank or more between flight

crewmembers had 21 percent more errors than flight crews in which the crewmembers were of

the same rank.

113

In a study of 249 airline pilots in the United Kingdom, nearly 40 percent of

first officers stated that they failed to communicate safety concerns to their captains on more

112

See KLM B-747, PH-BUF, and Pan Am B-747, N736PA, Collision at Tenerife Airport, Los Rodeos, Spain,

on 27 March 1977, Aircraft Accident Report (Spain: Subsecretaría de Aviación Civil, 1978).

113

R.A. Alkov and others, “The Effect of Trans-Cockpit Authority Gradient on Navy/Marine Helicopter

Mishaps,” Aviation, Space, and Environmental Medicine, vol. 63, no. 8 (1992), pp. 659-661.

NTSB Aircraft Accident Report

than one occasion for reasons that included a desire to avoid conflict and deference to the

captain’s experience and authority.

114

The poor CRM observed in this accident is similar to that which the NTSB found during

its investigation of the February 19, 1996, crash of Continental Airlines flight 1943 in Houston,

Texas.

115

During the event sequence leading to the accident, the captain rejected, without any

discussion, the first officer’s go-around request, and the first officer failed to challenge his

decision. The flight crew failed to complete multiple checklists and continued an unstabilized

approach below 500 feet

116

against company procedures to go around in such situations. As a

result of this accident, the NTSB issued Safety Recommendations A-97-05 and -06,

117

which

asked the FAA to require all POIs of 14 CFR Part 121 carriers to 1) ensure that the carriers

establish a policy and make it clear to their pilots that there will be no negative repercussions for

appropriate questioning in accordance with CRM techniques of another pilot’s decision or action

and 2) ensure that CRM programs provide pilots with training in recognizing the need for, and

practice in presenting, clear and unambiguous communications of flight-related concerns.

On October 30, 1998, the FAA issued AC 120-51C (currently AC 120-51E), which

emphasized the importance of all levels of management to support a safety culture that promotes

communication by encouraging appropriate questioning and that there will be no negative

repercussions for appropriate questioning of one pilot’s decision or action by another. Based on

the FAA’s lack of action to issue an FSIB, on February 23, 2000, the NTSB classified these

recommendations “Closed—Unacceptable Action.” Thirteen years after the FAA issued

AC 120-51C, the NTSB continues to investigate accidents in which one pilot does not question

the actions or decisions of another pilot.

To overcome large authority gradients, AC 120-51E suggests that assertiveness be

included in CRM training programs. FlightSafety’s CRM training materials for Empire Airlines’

flight crewmembers included two presentation slides that encouraged pilots to speak up with

“appropriate persistence until there is a clear resolution.” However, the accident first officer

could not recall what was trained during the 2-day CRM training course, including whether or

not assertiveness was addressed.

According to Empire Airlines’ ATR 42 pilot handbook, if a go-around maneuver is

required, the PF should call out “go around” and initiate the go-around procedures. During the

accident flight, the first officer instead asked the captain, “should I go around?,” rather than

directly asserting her concern about the unstabilized approach, even though she was aware that

company policy was to go around in the event of stick-shaker activation. The first officer

indicated that asking was her way of saying that she wanted to go around “without stepping on

toes.” Captains who had previously flown with the first officer stated that, although she did not

seem to have a problem standing up for something in the cockpit, she asked a lot of questions

when flying that were related to skills that she already knew.

114

J. Wheale, “Crew Coordination on the Flight Deck of Commercial Transport Aircraft,” Proceedings of the

Flight Operations Symposium, October 1983 (Dublin: Irish Air Line Pilots Association/Aer Lingus, 1983).

115

NTSB/AAR-97/01.

116

A ground proximity warning system alert continued below 200 feet above field elevation.

117

For more information about Safety Recommendations A-97-05 and -06, see section 1.18.2.4.

NTSB Aircraft Accident Report

The first officer stated that, after the captain responded “no” to her go-around inquiry, she

felt that he had a good reason for not wanting to go around and that she trusted that he was

making the right decisions. The first officer stated that, after the captain took control of the

airplane, she was still concerned with the approach and felt that she should have called again for

a go-around maneuver but that she did not know why she did not say anything. The NTSB

concludes that the first officer’s failure to assert herself to the captain and initiate a go-around

maneuver when she recognized the unstabilized approach likely resulted from the steep authority

gradient in the cockpit and the first officer’s minimal training on assertiveness; further, the

captain’s quick dismissal of the first officer’s go-around inquiry likely discouraged the first

officer from voicing her continued concerns and challenging the captain’s decision to continue

the unstabilized approach. The NTSB notes that the first officer’s CRM training did not include

any role-playing activities in which pilots could practice developing assertiveness skills. Practice

allows pilots to bridge the gap between their knowledge of assertiveness and the actions needed

in the cockpit to effectively be assertive.

118

The NTSB concludes that role-playing exercises are

essential for effective assertiveness training because such exercises provide flight crews with

opportunities for targeted practice of specific behaviors and feedback that a lecture-based

presentation format lacks. Therefore, the NTSB recommends that the FAA require that

role-playing or simulator-based exercises that teach first officers to assertively voice their

concerns and that teach captains to develop a leadership style that supports first officer

assertiveness be included as part of the already required CRM training for 14 CFR Part 121, 135,

and 91 subpart K pilots.

2.3.4 Go-Around Callout

Empire Airlines’ ATR 42 pilot handbook specified that the phrase “go around” is to be

stated out loud by the PF; however, the operating procedures did not provide any guidance

indicating that the PM could initiate the same action. Similarly, Empire Airlines’ flight training

materials contained no guidance that indicated whether either pilot could call for a go-around

maneuver, if necessary.

During the NTSB’s investigation of the February 18, 2007, Shuttle America runway

overrun accident,

119

the NTSB noted that it is critical to flight safety that either flight

crewmember be able to call for a go-around maneuver if either pilot believes that a landing

would be unsafe. As a result, the NTSB issued Safety Recommendation A-08-18,

120

which asked

the FAA to require operators to have a written policy emphasizing that either pilot can make a

go-around callout and that the proper response to such a callout is an immediate missed

approach. On March 1, 2010, the FAA issued SAFO 10005, “Go-Around Callout and Immediate

Response,” which recommended that operators provide such a written policy. However, the FAA

has not yet revised FAA Order 8900.1 to include guidance to POIs to have operators develop the

written policy and has not yet completed a survey to ensure that all operators have developed and

118

R.L. Beard, E. Salas, and C. Prince, “Enhancing Transfer of Training: Using Role Play to Foster Teamwork

in the Cockpit,” International Journal of Aviation Psychology, vol. 5, no. 2 (1995), pp. 131-143.

119

NTSB/AAR-08/01.

120

For more information about Safety Recommendation A-08-18, see section 1.18.2.2.

NTSB Aircraft Accident Report

implemented this policy. The NTSB encourages the FAA to take timely action in response to this

recommendation.

2.3.5 Training for Single and Multiple Abnormal Situations Occurring at Low

Altitude

As this accident shows, pilots can encounter rapidly changing conditions during

approach-to-landing operations. The accident flight crew encountered a flap asymmetry while

operating in moderate icing conditions. (Moderate icing conditions, although considered a

normal situation in the AFM, increase flight crew workload.) Because any distraction from

maintaining airplane controllability at low altitude can lead to dire consequences, pilots must be

familiarized with the rapid decision-making and maneuvering skills required to ensure safe

flight, especially when operating near the ground. During the accident flight, the captain

remained committed to continuing the approach, even though he was trained that (and SOPs

required that) a go-around maneuver should be performed and even after he received multiple

indications that the approach had become dangerously unstabilized.

After the accident, the captain explained that he “just wanted to land as soon as possible.”

The 1998 NASA study

121

found that the most common decision errors occurred when the flight

crew decided to “continue with the original plan of action in the face of cues that suggested

changing the course of action.” The NASA study stated that the cues that signal a problem are

not always clear and that a decision maker’s assessment of the situation may not keep pace with

conditions that deteriorate gradually. The NASA study also indicated that individuals have a

natural tendency to maintain their originally selected course of action until clear and

overwhelming evidence exists that the course of action should be changed. In addition, the study

noted that pilots under stress might not evaluate the consequences of various options.

On May 19, 2009, as a result of the NTSB’s investigation of the September 28, 2007,

American Airlines flight 1400 accident,

122

which involved multiple, simultaneous abnormal

situations, the NTSB issued Safety Recommendations A-09-24 and -25,

123

which asked the FAA

to establish best practices for conducting both single and multiple emergency and abnormal

situations training and to require that these best practices be incorporated into all operators’

approved training programs. Although the FAA has made progress by publishing its January 12,

2009, NPRM, “Qualification, Service, and Use of Crewmembers and Aircraft Dispatchers,”

124

which included scenario-based training for all 14 CFR Part 121 carriers, the NTSB is concerned

that, without a guide to best practices, inspectors and operators will have difficulty creating

effective programs. The FAA has not yet completed a review of the regulatory and policy

framework to determine whether additional guidance or requirements are necessary.

121

J. Orasanu, L. Martin, and J. Davison, Errors in Aviation Decision Making: Bad Decisions or Bad Luck?,

NASA Ames Research Center, Moffett Field, California. Presented to the Fourth Conference on Naturalistic

Decision Making, Warrenton, Virginia, May 29-31, 1998.

122

NTSB/AAR-09/03.

123

For more information about Safety Recommendations A-09-24 and -25, see section 1.18.2.1.

124

74 FR 1280. (The FAA subsequently began work on an NPRM supplement, which was submitted to the

Office of Management and Budget (OMB) for review on November 29, 2010.

As of the date of this report, OMB is

continuing to review this supplemental NPRM with no projected completion date.)

NTSB Aircraft Accident Report

The NTSB recognizes that operators have limited time to allocate to training. However,

this accident and others show that pilots could benefit from opportunities to develop the skills

and decision-making abilities needed during single and multiple abnormal situations; such

training can present pilots with competing task demands and increased workloads and can

demonstrate in a safe environment the possible consequences of not taking timely, appropriate

corrective action. This accident illustrates that these skills are most important at low altitudes

when rapid, accurate assessment of abnormal situations and appropriate prioritization of tasks are

the most critical. Thus, the NTSB concludes that the establishment of best practices for

conducting both single and multiple emergency and abnormal situations training needs to include

training for the occurrence of these situations at low altitudes because low-altitude scenarios

require rapid, accurate assessment of abnormal situations and appropriate prioritization of tasks.

The circumstances of this accident reinforce the need for the actions recommended in Safety

Recommendations A-09-24 and -25, and the NTSB encourages the FAA to take timely action in

response to these recommendations.

2.3.6 Fatigue Considerations

The NTSB evaluated a number of criteria

125

to determine the extent to which fatigue

impacted the flight crew’s performance during the accident sequence including circadian factors,

sleep length, and time awake.

The accident occurred about 0437, and the flight crewmembers were awake in opposition

of their normal circadian rhythm. For those entrained on a diurnal schedule, the primary

circadian trough is about midnight to 0600, with the window of circadian low generally

occurring between about 0300 and 0500. The aeromedical research community has identified

strategies that flight crews can use to lessen the adverse effects of fatigue during required

overnight operations. These strategies include avoiding morning sunlight and staying indoors,

ensuring the daytime sleep environment is dark and cool, wearing eye masks and earplugs during

daytime sleep, implementing good sleep habits, taking a short nap before reporting for duty,

allowing exposure to at least 2 hours of sunlight or bright artificial light in the late afternoon or

early evening after waking, strategically using caffeine, and getting regular exercise.

126

Both

flight crewmembers took some actions before the accident to reduce the likelihood of

performance decrements associated with being awake during the nighttime hours.

For example, during the week before the accident flight (January 20 through 23), the first

officer had a scheduled trip in which she was on duty during the night hours beginning about

0045. She was off duty from the afternoon of January 23 through the evening of January 26.

Empire Airlines paid her expenses to spend the weekend in Midland, Texas, rather than

commuting back to her home in Washington for a short turnaround.

127

The first officer stated

125

Information regarding the NTSB’s methodology for investigating operator fatigue is available online at

<http://www.ntsb.gov/info/fatigue_checklist_V%202_0.pdf>.

126

M. Caldwell and others, “Fatigue Countermeasures in Aviation,” Aviation, Space, and Environmental

Medicine, vol. 80, no. 1 (2009), pp. 29-59.

127

The flight crew was based in Salt Lake City, which was considered to be a “floater” base. The flight

crewmembers would travel to various cities based on the needs of the company, which paid their expenses over the

weekend.

NTSB Aircraft Accident Report

that because she had acclimated herself to sleeping during the days and being awake at night for

the previous trip, she remained on that sleep/wake schedule during her off-duty days rather than

resetting her circadian rhythm to a diurnal pattern of being awake during the day and sleeping at

night. During the 3 days before the accident, the first officer indicated that she went to bed about

0600 and awoke between about 1400 and 1500, providing a sleep opportunity each night of

between 8 and 9 hours.

128

Based on these times, the first officer went to bed before sunrise, thus

avoiding morning sunlight, and awoke before sunset, allowing exposure to bright daylight before

reporting for duty, which would have contributed to her ability to shift her circadian clock.

129

Research suggests that adjustment to night activity is possible, and, under ideal conditions, the

adjustment occurs about 1 hour per day.

130

However, a NASA study examining pilots in

overnight cargo operations found that the circadian clock of pilots did not shift completely as

measured by temperature minimums;

131

the circadian low was delayed about 3 hours after 5 days

of night flying.

132

Although the first officer’s efforts to adapt to an overnight schedule suggest an

individual awareness of the need to mitigate fatigue hazards associated with overnight

operations, and the first officer was possibly able to delay her window of circadian low to some

degree, it is unlikely, based on the available research data, that her efforts would have completely

mitigated the circadian effects of fatigue associated with the time at which the accident occurred.

Sleep needs vary by individual. In postaccident interviews, the first officer stated that she

needed 7 hours of sleep to feel rested. In the 72 hours before the accident, she had 25.5 hours of

sleep opportunity (8.5 hours, 8 hours, and 9 hours in the previous three nights). She indicated

that she felt rested on the evening of the accident. The first officer’s sleep schedule during her

previous trip is unknown. The NASA study examining pilots in overnight cargo operations found

that, while on duty, 54 percent slept over 1 hour less per 24 hours than they did on pre-trip days,

29 percent slept over 2 hours less per 24 hours than they did on pre-trip days, and 15 percent

averaged more sleep per 24 hours than they did on pre-duty days.

133

Post-duty days allowed

pilots to recuperate some of their lost sleep. Specifically, the pilots in the study slept 41 minutes

longer per 24 hours during post-duty days than pre-trip days and 1.9 hours longer per 24 hours

during post-duty days than duty days. Although the first officer had sleep opportunities in excess

of the hours of sleep she needed to feel rested during her off-duty days and had the opportunity

to recuperate any sleep debt accrued, it is unknown as to how much sleep she actually received.

Therefore, it cannot be determined whether the first officer had accumulated a sleep debt during

the week before the accident.

128

The first officer stated that she needed 7 hours of sleep per night to feel rested.

129

Night conditions would have prevailed at the time of the first officer’s reported bedtime, which was more

than 1 hour before sunrise in Midland, Texas. For the week before the accident, sunrise ranged from about 0745 to

0748, and sunset ranged from about 1810 to 1817.

130

R. Wever, “Phase Shifts of Circadian Rhythms due to Shifts of Artificial Zeitgebers,” Chronobiologia,

vol. 7 (1980), pp. 303-327.

131

Core body temperature minimums typically occur in the circadian trough and can be associated with sleep

tendency. Caldwell and others, p. 42.

132

P.H. Gander and others, “Flight Crew Fatigue IV: Overnight Cargo Operations,” Aviation, Space, and

Environmental Medicine, vol. 69, no. 9 (1998), pp. B26-B36.

133

Gander and others, pp. B28-B29.

NTSB Aircraft Accident Report

At the time of the accident, the first officer had been awake about 13.5 hours. Research

into quantifying performance impairment associated with sustained wakefulness found that

performance remains relatively stable throughout the time that coincides with a normal waking

day, but that prolonged wakefulness of 17 hours can result in measurable performance

impairment (comparable to having a blood alcohol concentration of 0.05 percent).

134

The first

officer did not have such prolonged or otherwise excessive periods of wakefulness before the

accident.

The captain also took steps to minimize the effects of fatigue due to the circadian clock.

On the night before the accident trip, the captain went to sleep about 2200 but, to prepare himself

for the upcoming night flight, awoke about 0400 on January 26, 2009. Later that afternoon, he

napped from about 1100 to 1630, providing a 5.5-hour sleep opportunity. The captain stated that

he needed 6 to 8 hours of sleep each night to feel rested. In the 72 hours before the accident, the

captain had 22 hours of sleep opportunity (10 hours, 6 hours, and 5.5 hours). The captain stated

that he felt rested on the evening of the accident. Although research suggests that appropriately

placed naps can improve alertness for up to 24 hours,

135

because the accident occurred during the

window of circadian low, the captain was likely experiencing some fatigue at the time of the

accident. In addition, if the captain required 8 hours of sleep per night to feel rested, he may have

been experiencing as much as 4.5 hours of cumulative sleep loss on the night of the accident.

Research suggests that as few as 2 hours of sleep loss can lead to reduced performance and

alertness.

136

At the time of the accident, the captain had been awake about 12 hours, which

would not be considered excessive.

2.3.6.1 Role of Fatigue in Flight Crew Performance

The negative effects of fatigue on human performance have been demonstrated in

scientific research and accident and incident investigations.

137

These effects include slowed

response time, reduced vigilance, and poor decision making. The flight crew’s failure to monitor

134

(a) D. Dawson and K. Reid, “Fatigue, Alcohol and Performance Impairment,” Nature, vol. 388, no. 6639

(1997), p. 235. (b) N. Lamond and D. Dawson, “Quantifying the Performance Impairment Associated with

Sustained Wakefulness” (Woodville, South Australia: The Centre for Sleep Research, The Queen Elizabeth

Hospital, 1998). <http://cf.alpa.org/internet/projects/ftdt/backgr/Daw_Lam.html> (accessed March 21, 2011).

135

D.F. Dinges and others, “Temporal Placement of a Nap for Alertness: Contributions of Circadian Phase and

Prior Wakefulness,” Sleep, vol. 10, no. 4 (1987), pp. 313-329.

136

R. T. Carskadon, “Sleep Restriction,” ed., Sleep, Sleepiness and Performance, T. H. Monk, ed. (Chichester,

England: John Wiley & Sons, 1991), pp. 155–167.

137

For the scientific research, see J.A. Caldwell, “Fatigue in the Aviation Environment: An Overview of the

Causes and Effects as Well as Recommended Countermeasures,” Aviation, Space, and Environmental Medicine,

vol. 68 (1997), pp 932-938; G.P. Kruger, “Sustained Work, Fatigue, Sleep Loss, and Performance: A Review of the

Issues,” Work and Stress, vol. 3 (1989), pp 129-141; and F.H. Previc and others, “The Effects of Sleep Deprivation

on Flight Performance, Instrument Scanning, and Physiological Arousal in Pilots,” The International Journal of

Aviation Psychology, vol. 19, no. 4 (2009), pp. 326-346. For the accident investigations, see, for example, Crash

During Attempted Go-Around After Landing East Coast Jets Flight 81 Hawker Beechcraft Corporation 125-800A,

N818MV Owatonna, Minnesota July 31, 2008, Aircraft Accident Report NTSB/AAR-11/01 (Washington, DC:

National Transportation Safety Board, 2011). <http://www.ntsb.gov/Publictn/2011/AAR-11-01.pdf>; Collision with

Trees and Crash Short of Runway, Corporate Airlines Flight 5966, British Aerospace BAE-J3201, N875JX,

Kirksville, Missouri, October 19, 2004, Aircraft Accident Report NTSB/AAR-06/01 (Washington, DC: National

Transportation Safety Board, 2006). <http://www.ntsb.gov/publictn/2006/AAR0601.pdf>.

NTSB Aircraft Accident Report

the airplane’s airspeed following the flap anomaly could be consistent with the known effects of

fatigue; however, this type of error has also been made by flight crews who were not fatigued.

Throughout the flight, the first officer demonstrated good performance, including

correcting the captain when setting up the approach, questioning the meaning of the Mu readings

provided by ATC, and identifying the need to execute a go-around maneuver after the stall

warning activated. Her failure to monitor airspeed and maintain a stabilized approach did not

occur until after the flap anomaly, while the captain was performing a nonstandard

troubleshooting procedure. The first officer queried the captain as to whether she should go

around, and he told her “no.” The first officer was flying an airplane in which she had little

experience and in conditions that she had never before experienced. As a result, the first officer

relied on the captain’s leadership and experience for guidance, and his instructions to continue

the unstabilized approach contradicted SOPs and likely contributed to her failure to assert her

concerns further. In addition, the captain failed to delegate duties in the abnormal situation, and

his nonstandard approach to address the flap anomaly likely distracted her from her PF duties.

The NTSB concludes that, although the risk for fatigue existed at the time of the accident due to

the window of circadian low, the first officer took steps to mitigate the effects of fatigue, and her

errors during the flight can be explained by her lack of experience in both the airplane and in

icing conditions along with the distraction caused by the captain’s nonstandard response to the

flap anomaly.

Unlike the first officer, the captain made a number of errors throughout the flight,

including briefing the incorrect approach airspeeds, failing to monitor the first officer’s

approach, and performing a nonstandard and nontrained procedure in response to the flap

anomaly. The captain’s decision not to follow SOPs for checklist usage and the unstabilized

approach might be explained through his actions demonstrated on previous flights (another pilot

had indicated that the captain occasionally cut corners, seemed rushed, and was less thorough on

briefings than other captains) and the stress associated with the situation. However, the captain’s

repeated errors and decision to forego procedures might also reflect degradation from fatigue as a

result of performing during a time when he was normally asleep, as well as performing with a

sleep debt. The NTSB concludes that, due to the early morning hour of the accident flight and

the captain’s cumulative sleep debt, fatigue likely degraded the captain’s performance. The

NTSB notes that careful adherence to SOPs, including strict adherence to accepted practices of

stabilized approach criteria, non-normal procedures, and CRM, can help flight crews mitigate the

degrading effects of fatigue. The NTSB further notes that neither the captain nor the first officer

received specific training from their company addressing the hazards of fatigue or fatigue

mitigation, although fatigue was briefly addressed during CRM training.

On September 10, 2010, the FAA addressed some aspects of fatigue by issuing an NPRM

titled “14 CFR Parts 117 and 121: Flightcrew Member Duty and Rest Requirements,” which

proposes to amend Part 121 and establish Part 117 to create a single set of flight time limitations,

duty period limits, and rest requirements for Part 121 operations. The NPRM also proposes

requiring a fatigue education and training program for all flight crewmembers, employees

involved in operational control and scheduling of flight crewmembers, and personnel having

management oversight of these areas. The FAA is currently drafting AC 120-FT, “Fatigue

Training,” that will present guidelines for developing and implementing a fatigue training

program. The draft guidance outlines a training program that includes, in part, a review of the

NTSB Aircraft Accident Report

basics of fatigue, including an overview of circadian rhythms, the causes of fatigue, the effect of

fatigue on performance, fatigue countermeasures and mitigation, and lifestyle influences on

fatigue.

On November 15, 2010, the NTSB commented on the NPRM and stated, “if adopted, the

proposed rule will provide substantial benefits towards reducing the hazards associated with

flight crew fatigue in Part 121 operations.” The NTSB also stated its strong support for the

requirement of a fatigue education and training program, noting that fatigue education is among

the foundational elements of an effective fatigue management system.

The NTSB is encouraged by the FAA’s response to fatigue mitigation for Part 121

operations based on the NPRM. In recent years, a number of accidents involving Part 121

operators have occurred in which fatigue was causal, contributory, or identified as a safety issue.

The NTSB believes that finalizing the rule in a timely fashion will provide substantial benefits

towards reducing the hazards associated with flight crew fatigue in Part 121 operations and is

critical to aviation safety.

2.4 Operations Issues

2.4.1 Dispatch into Freezing Drizzle Conditions

The flight release for the accident flight contained weather information that indicated that

IFR conditions with light freezing drizzle and mist were reported at LBB and that such

conditions were forecast to continue beyond the flight’s estimated time of arrival. During the

flight’s descent toward LBB, the controller informed the flight crew that light freezing drizzle

and mist were reported at the airport.

Although SLD conditions like freezing rain or freezing drizzle are outside the ATR 42’s

certification icing envelope as specified in 14 CFR Part 25, Appendix C, the FAA does not

specifically prohibit operators from dispatching or operating airplanes in such weather

conditions. Instead, the FAA requires that operators adhere to any AFM limitations for the

airplanes that they operate. The NTSB has long been concerned about such operations and issued

numerous safety recommendations in the 1980s

138

and the 1990s

139

related to the issue. In

response to these recommendations, the FAA issued an NPRM.

140

For the ATR 42, the AFM states that a flight crew must immediately exit severe icing

conditions when they are encountered. According to Empire Airlines’ GOM, “when light

freezing rain, light or moderate freezing drizzle, or light, moderate, or heavy snow is falling,

aircraft may land.” The GOM also stated that, “when light freezing rain, light to moderate

freezing drizzle, or light, moderate snow is falling, [an ATR] aircraft may take off, provided it is

prepared in accordance with approved deicing procedures.”

138

NTSB/SR-81/01.

139

NTSB/AAR-96/01.

140

For more information about the NPRM, see section 2.4.1.2.

NTSB Aircraft Accident Report

After the accident, on February 27, 2009, Empire Airlines issued a flight operations

bulletin to supersede the icing information in the GOM and to inform flight crews and flight

followers that freezing rain and freezing drizzle are not covered by the icing certification

envelope and that takeoff or landing operations in known or reported freezing rain or freezing

drizzle of any intensity are prohibited. An informal review of other 14 CFR Part 121 and 135

operators’ icing guidance revealed that the operators’ guidance for dispatching their

turbine-powered, pneumatic deice boot-equipped airplanes into freezing rain and freezing drizzle

conditions varied widely; some operators prohibited dispatch into such conditions, and others

permitted it.

As this accident shows, moderate icing conditions due to light freezing drizzle can

increase a flight crew’s workload and degrade the performance of the airplane. Although Empire

Airlines and the other FedEx feeder operators developed “no-go” weather guidance (after the

accident) that prohibits takeoff or landing operations in known or reported freezing rain or

freezing drizzle, these actions were not required by the FAA, and other operators may not adopt

such safety measures. The NTSB concludes that dispatching and operating an airplane in known

icing conditions for which the airplane is not certificated and has not demonstrated the ability to

operate safely has the potential to reduce or eliminate safety margins. Therefore, the NTSB

recommends that the FAA prohibit all 14 CFR Part 121, 135, and 91 subpart K operators of

pneumatic deice boot-equipped airplanes from dispatching or deliberately operating these

airplanes in known freezing rain or freezing drizzle of any intensity, unless the airplane

manufacturer has demonstrated that the airplane model can safely operate in those conditions.

2.4.1.1 Flight Crewmember and Dispatcher Training Related to Icing Conditions

According to Empire Airlines’ director of operations, the GOM information in place

before the accident (which permitted dispatching and operating airplanes in light freezing rain;

light or moderate freezing drizzle; or light, moderate, or heavy snow) was developed based on

information from the ADP holdover tables. However, the NTSB notes that holdover tables are

references that are used for ground deicing operations and are not intended to be used to address

the in-flight icing environment.

141

Interviews with Empire Airlines’ flight followers and the dispatch manager revealed that

their perceptions about dispatching into freezing drizzle conditions were based on this GOM and

ADP information. However, such guidance based on ground deicing operations did not

emphasize the potential dangers of in-flight encounters with freezing rain and freezing drizzle

that are described in airplane-specific flight operations publications. For example, the AFM for

the airplane (which is a flight publication provided by ATR) contained the following warning:

Severe icing may result from environmental conditions outside of those for which

the airplane is certificated. Flight in freezing rain, freezing drizzle, or mixed icing

141

In the section of the March 2011 edition of Cold Weather Operations: Be Prepared for Icing that discusses

ground deicing, ATR states, in part, the following: “IMPORTANT: The fluids used during ground de/anti-icing is

not intended for and does not provide protection during the flight. Before take-off, pilots must be aware of potential

in-flight severe icing conditions…. If such conditions exist, take-off must be delayed”

NTSB Aircraft Accident Report

conditions (supercooled liquid water and ice crystals) may result in ice buildup on

protected surfaces exceeding the capabilities of the ice protection system.

Also, ATR’s Cold Weather Operations publication

142

stated that in-flight icing is a major

concern for commuter airplanes (which are typically pneumatic deice boot-equipped airplanes)

in particular because they fly at altitudes where icing conditions are most likely to occur. Neither

of these flight publications was used in developing dispatch guidance.

Based on information obtained during interviews with flight crewmembers and dispatch

personnel, the NTSB found that many were not sufficiently aware of the dangers of operating in

freezing drizzle precipitation and lacked a thorough understanding of the weather phenomena

associated with SLD conditions. Although on March 16, 2010, the FAA issued SAFO 10006,

which referenced several of the NTSB’s safety recommendations that resulted from its

investigation of the October 31, 1994, accident in Roselawn, Indiana,

143

to encourage all

operators to review and amend, if necessary, their flight crewmember and dispatcher training

programs to ensure that their programs address SLD icing conditions, the actions recommended

in a SAFO are not mandatory. Also, as noted previously, other operators’ dispatch guidance

varied, and Empire Airlines erroneously relied on ground deicing guidance as the basis for

dispatch decisions, which resulted in assumptions that led to the dispatch of flights into freezing

drizzle without adequate consideration for the in-flight hazards associated with the conditions.

The NTSB is concerned that disparity exists between operators’ and airplane

manufacturers’ guidance materials and between the guidance materials of operators of similar

airplanes regarding flight operations in SLD conditions, particularly after the NTSB issued

numerous safety recommendations on the topic resulting from its investigation of the October 31,

1994, accident in Roselawn, Indiana.

144

Dispatch personnel like Empire Airlines’ flight

followers who authorize each flight release, provide preflight weather products, and perform

flight following share an integral role with flight crews in ensuring that each flight safely reaches

its destination. The NTSB concludes that, to most effectively ensure the safety of flight

operations in icing conditions, pilots, dispatchers, and flight followers must understand how the

dangers of freezing drizzle and freezing rain can affect their airplanes and must understand the

differences between ground deicing considerations and in-flight icing operations. Therefore, the

NTSB recommends that the FAA review the approved pilot, dispatcher, and flight follower

training programs and procedures for all 14 CFR Part 121, 135, and 91 subpart K operators and

require revisions to the programs and procedures, as necessary, to include standardized training

and aircraft-specific information to educate pilots, dispatchers, and flight followers of the

dangers of flight operations in freezing precipitation and of the differences between ground

deicing considerations and in-flight icing operations.

142

Cold Weather Operations: Be Prepared for Icing (2008).

143

NTSB/AAR-96/01.

144

NTSB/AAR-96/01.

NTSB Aircraft Accident Report

2.4.1.2 Proposed Rulemaking for Airplane Certification Requirements for Flight in

Icing Conditions

In response to several NTSB safety recommendations, on June 29, 2010, the FAA issued

an NPRM

145

intended to improve the safety of transport-category airplanes operating in SLD,

mixed-phase, and ice-crystal icing conditions. The NPRM included proposals for expanding the

certification icing environment for newly certificated transport-category airplanes to include

freezing rain and freezing drizzle and for requiring airplanes most affected by SLD icing

conditions to meet certain safety standards in the expanded certification icing environment,

including additional airplane performance and handling requirements.

On August 27, 2010, in comments submitted to the docket for this NPRM, the NTSB

stated that it was pleased that the NPRM proposed to add new sections to 14 CFR Part 25.

Specifically, the NTSB noted section 25.1420, which would require aircraft manufacturers to

evaluate airplane operation in the SLD environment, and Appendix O, Part I, which would

expand the certification icing environment to include freezing rain and freezing drizzle by using

four separate droplet-size distributions. Additionally, the NTSB stated that it was pleased to see

the inclusion of Appendix O, Part II, which states the following:

The most critical ice accretion in terms of airplane performance and handling

qualities for each flight phase must be used to show compliance with the

applicable airplane performance and handling qualities requirements for icing

conditions contained in Subpart B of this part. Applicants must demonstrate that

the full range of atmospheric icing conditions specified in Part I of this appendix

have been considered, including drop diameter distributions, liquid water content,

and temperature appropriate to the flight conditions.

Although the NTSB finds that several aspects of the FAA’s NPRM are essential for

improving the safety of flight for airplanes operating in icing conditions, the NTSB is

disappointed that the proposed rule applies only to newly certificated aircraft (thus, the accident

airplane would not be subject to the new rules) and that it excludes certain aircraft (such as those

with a maximum takeoff weight of more than 60,000 lbs or irreversible flight controls) from

compliance with the new safety standards. In its response, the NTSB noted that SLD conditions

can present a danger for the current fleet of aircraft; therefore, the proposed rule should be

expanded beyond newly certificated airplanes to include all deice boot-equipped airplanes

currently in service that are certificated for flight in icing conditions.

2.4.2 Training Program and Manual Discrepancies

Empire Airlines personnel indicated that, during basic indoctrination training, pilots

reviewed the ADP, FAA’s meteorology handbook, and icing videos. Empire Airlines personnel

also indicated that, during training at FlightSafety, pilots reviewed the ATR Cold Weather

Operations publication

146

and reviewed meteorological conditions likely to cause freezing

145

75 FR 37311-37339.

146

Cold Weather Operations: Be Prepared for Icing (2008).

NTSB Aircraft Accident Report

drizzle and freezing rain. However, Empire Airlines’ FTM included a document that indicated

that Empire Airlines (not FlightSafety) was required to provide the required meteorology

training, including the ATR publication, as part of the meteorology module. Interviews with

FlightSafety instructors revealed that they were not providing or using the publication during

training. The flight crew did not recall receiving this document.

Although Empire Airlines implemented special emphasis icing training for all ATR pilots

after the accident, the NTSB had concerns that the FAA did not detect these training program

discrepancies before the accident. Spokane FSDO personnel managed their oversight of Empire

Airlines through the FAA’s ATOS program. A review of the ATOS program table,

“System/Subsystem/Element Chart—Operations and Cabin Safety Elements,” revealed that, as

part of the program, the FAA was required to provide oversight of an airline’s FTM content and

flight crewmember training. However, the inspectors who followed the ATOS program did not

detect that Empire Airlines was not providing the required meteorology training, including the

ATR publication, as part of the meteorology training module. Upon learning of this, the FAA

contacted the certificate management offices of other ATR operators and determined that this

was an oversight problem in the inspection process at Empire Airlines and not a systemic issue

with ATOS. Additionally, the FAA took corrective action to assure that changes to Empire

Airlines’ FTM were initiated to assure that the required meteorological training was being

conducted by the operator.

2.5 Survival Aspects

According to dispatch logs, the accident alarm sounded at 0438:12. ARFF personnel who

were dispatched to the scene reported that they proceeded on the taxiways with caution because

of the ice. The ARFF response vehicles arrived on scene without difficulty at 0442:31, and

ARFF personnel immediately began fire suppression and contained the fire (with the exception

of a few interior cargo fires) within minutes. Thus, the NTSB concludes that the presence of ice

on the taxiways between the ARFF station and the accident site minimally increased the response

time; however, the emergency response was timely and effective in suppressing the fire.

2.5.1 Lack of Occupant and Hazardous Materials Information Available to First

Responders

The first officer opened the cockpit escape hatch without difficulty, but the flight crew

decided not to use it for egress because of the fire’s proximity. The captain and the first officer

opened the forward cargo door without difficulty and exited the airplane. The first officer stated

that, before exiting the airplane, the captain attempted to contact the ATCT using the airplane’s

radio but that the radio was inoperative (the CVR did not capture this attempt). The captain

stated that, after exiting the airplane, he called Empire Airlines’ dispatch and reported the

accident. Both the captain and the first officer then left the scene and ran to the FedEx hangar

before the first responders arrived. A FedEx employee subsequently notified the Lubbock Fire

Department and EMS personnel when they arrived at LBB that the flight crew was at the FedEx

hangar; however, by that time, the ARFF responders were already on scene and looking for

survivors.

NTSB Aircraft Accident Report

As a result of the confusion regarding the flight crewmembers’ whereabouts, at least two

ARFF personnel unnecessarily accessed the airplane to look inside the cockpit and cabin area,

and other personnel also unnecessarily interrupted their fire suppression activities to participate

in the search. After the arriving fire department personnel conveyed to ARFF personnel that the

flight crew was safe, the search was stopped but then was resumed briefly after the ATCT

controller relayed information (which was apparently old) that a pilot was walking around on the

ramp looking for assistance. In addition, the ARFF personnel were initially unaware of the

airplane’s HAZMAT cargo. Operators that transport HAZMAT cargo are required to make

specific HAZMAT cargo information available during an emergency response, and Empire

Airlines’ procedures complied with the requirements. However, the requirements do not address

situations in which the flight crew may be able to provide more timely, general information in

advance of the operator’s delivery of the required documents. The ARFF captain indicated that,

if they had the information, the ARFF response to the scene would have been similar with the

exception that all responding firefighters would have worn full personal protective equipment.

Although the flight crewmembers acted appropriately by ensuring their own safety

following the accident, the situation highlights that a lack of information from the flight crew can

hinder the response team’s ability to safely and most efficiently prioritize their on-scene

activities. The NTSB notes that a similar situation occurred on June 28, 2008, during a cargo

airplane ground fire in which the flight crew evacuated the airplane and left the scene before first

responders arrived, but no information about their safe egress was immediately provided to first

responders. In this accident, a firefighter unnecessarily entered the burning wreckage to look for

survivors and encountered a dangerous situation in the cockpit, including heavy smoke in which

he lost his radio, confined spaces that pulled off some of his equipment, and a “wall of fire”

behind the cockpit door. The firefighter indicated that, in hindsight, he should not have entered

the cockpit but that he did it because he did not know if anybody was in there and because he

thought that he could put the fire out.

147

The ARFF station was notified of the accident when the ATCT controller lifted the

direct-line radio receiver in the tower, resulting in the automatic sounding of a tone in the station.

According to procedure, following the tone, the ATCT controller can provide information about

the level of the alert, the type of aircraft, the nature of the emergency, the runway to be used, and

any other information as time permits. In the case of the accident flight, other than the accident

location, the only information about the flight that the ATCT controller could convey was that

the airplane was an ATR. ARFF personnel did not know that it was a cargo airplane and that

they could expect two or three occupants until they began suppressing the fire and saw the FedEx

logo on airplane’s tail. Further, as previously discussed, the ARFF responders were initially

unaware of the airplane’s HAZMAT cargo.

The NTSB concludes that timely information from the flight crew about the safety of the

airplane occupants and the presence of on-board HAZMAT cargo would improve the safety and

efficiency of the emergency response. Further, the NTSB concludes that, because flight crews

cannot always immediately communicate with ATC after an accident, it is important that another

147

See the Survival Factors Group Chairman’s Factual Report in the public docket for Ground Fire Aboard

Cargo Airplane, ABX Air Flight 1611, Boeing 767

‐

200, N799AX, San Francisco, California, June 28, 2008, Aircraft

Accident Summary Report NTSB/AAR-09/04/SUM (Washington, DC: National Transportation Safety Board,

2009). <http://www.ntsb.gov/Dockets/Aviation/DCA08MA076/407654.pdf>.

NTSB Aircraft Accident Report

method be developed to communicate information, such as the number of occupants on board

and the presence of hazardous materials, to ARFF personnel upon initial notification of an

accident. Therefore, the NTSB recommends that the FAA develop a method to quickly

communicate information regarding the number of persons on board and the presence of

hazardous materials to emergency responders when airport emergency response or search and

rescue is activated.

2.5.2 Inoperative Emergency Response and Mutual Aid Gate

The airport’s emergency plan indicated that units that respond from outside of the airport

should “respond to Gate 6 and Gate 48 or at a location prescribed by the command post.”

However, the emergency response unit that arrived at Gate 48 could not access the airport at that

location because the gate was inoperative due to ice in its operating mechanism. The remaining

response units were able to divert to another gate for airport access, which minimally delayed

their response time.

The LBB snow and ice removal plan did not address checking or ensuring the operability

of the airport gates. A survey of several other 14 CFR Part 139 airports with operations affected

by winter weather revealed that none of the airports’ snow and ice operations plans addressed

checking or ensuring the operability of emergency response and mutual aid gates. Also,

AC 150/5200-30C did not address the issue.

The NTSB notes that airport operations personnel are already required by airport snow

and ice removal plans to conduct periodic inspections of the routes used by emergency response

vehicles. Therefore, the checking of mutual aid gates for operability could easily be added as an

item to be performed during these inspections. Further, following the accident, LBB voluntarily

added the checking of mutual aid gates for operability to its snow and ice operations plan. The

NTSB concludes that an iced and inoperable mutual aid gate could extend the response time of

mutual aid, which could delay the delivery of medical attention to accident survivors and result

in further fire damage to property. Therefore, the NTSB recommends that the FAA amend

AC 150/5200-30C to include guidance on monitoring and ensuring the operability of emergency

response and mutual aid gates during winter operations.

2.6 Airplane Equipment Options

2.6.1 Aircraft Performance Monitoring

ATR developed the APM system, which has been installed on new production ATR 42

and ATR 72 airplanes since November 2005, to enhance a flight crew’s ability to detect the

effects of severe icing conditions on the airplane. The APM is a condition-specific, low-airspeed

alerting system because it is designed to provide specific alerts in icing conditions only (level 2

or 3 icing protection must be engaged or an icing signal must be provided by the icing detector).

The APM calculates and compares the airplane’s actual performance (with respect to airspeed

and drag) with its expected performance, computes the actual minimum icing and severe icing

airspeeds, and provides alerts to the flight crew when low airspeed or performance degradation is

detected. These alerts prompt the flight crew to perform the specified QRH procedures

NTSB Aircraft Accident Report

(depending upon the type of alert) to respond to the effects of icing on the airplane’s

performance.

Although the APM is intended to help flight crews recognize and respond to performance

degradations developed while operating in severe icing encounters, the APM alerts are an

important safety enhancement for flights operating even in moderate icing conditions, like the

accident flight. For example, had the accident airplane been equipped with an APM, the blue

“CRUISE SPEED LOW” annunciator lights would have illuminated about 70 seconds before the

flap anomaly, and the amber “DEGRADED PERF” annunciator lights and a single chime would

have activated about 40 seconds before the flap anomaly. An APM would not have provided any

further alerts after the flight crew selected flap extension because the system requires a flaps-up

configuration.

According to ATR, the appropriate flight crew response to the “DEGRADED PERF”

alert includes maintaining red-bug airspeed plus 10 kts. The NTSB notes that, although the

accident flight’s hazardous low-airspeed situation resulted from the flight crew’s inadequate

airspeed management rather than the performance degradation from the moderate icing

encounter, had the accident airplane been equipped with an APM, the APM would have provided

icing-related alerts that may have alerted the flight crew of the need to be more attentive to

airspeed management during the approach. Thus, the NTSB concludes that the APM system is a

valuable low-airspeed alerting tool that can enhance safety in all icing conditions.

APM retrofits for older ATR 42 (like the accident airplane) and ATR 72 airplanes have

been available since June 2006, and, effective August 24, 2009, EASA required that all

EASA-participating operators (operators in the European Community, its Member States, and

the European third countries) of ATR 42 and 72 airplanes not originally equipped with the APM

system retrofit their airplanes with the system within 72 months or by the second “C” check,

whichever occurs first. The FAA did not take similar action. As a result, APM system retrofits

are optional for U.S. operators.

In a January 14, 2011, letter responding to a query from NTSB investigators, the FAA

stated that its decision not to require APM retrofits for U.S. operators was based, in part, on its

finding that, in five of the ten icing-related events that were cited by EASA when mandating the

APM, the flight crewmembers were aware they were in severe icing conditions but did not

follow mandatory operating limitations. Therefore, the FAA reasoned that although the APM

would have provided alerts in four of these five cases, “it cannot be determined if the flight crew

would have acted any differently in response to an APM alert than they did to observing the

severe icing cues.” The NTSB notes that, in times of high workload, an alert even a few seconds

earlier that an ice accretion has crossed the line from “normal” to severe would be beneficial

because it would allow the flight crew to take immediate action. Because the APM warns of

performance degradations when they begin to occur, even if the autopilot is on and the crew is

unable to detect degradation in performance, the APM can provide warnings to flight crews that

a hazardous low-airspeed situation is developing well before the icing conditions become severe.

The NTSB does not concur with the FAA’s position that APMs would not help flight crews in

icing conditions and believes that the FAA should reexamine its decision not to require APM

retrofits for U.S. operators. Therefore, the NTSB recommends that the FAA require all operators

NTSB Aircraft Accident Report

of ATR 42- and ATR 72-series airplanes to retrofit the airplanes with an APM system if they are

not already so equipped.

2.6.2 Flap Malfunction or Asymmetry Light

During the accident flight, the flap asymmetry was not quickly or accurately identified by

either flight crewmember, despite several available cues, including the cockpit flap position

indicator, the exterior wing flap position fairings, the left control wheel input by the autopilot

(and the resulting 5° heading change of the airplane), the airplane’s failure to decelerate as

expected, and the “RETRIM ROLL L WING DN” and “AILERON MISTRIM” messages on the

ADU. Although it is possible that the flight crewmembers did not detect all of the cues (both the

captain and the first officer stated that they did not see any messages on the ADU), the cues they

did observe were not rapidly interpreted as indicative of a flap problem.

The NTSB notes that the ATR 42-320 (the accident airplane model) is one of only a few

models of transport-category airplanes certificated without a flap malfunction light (either a light

to indicate any flap malfunction or a specific asymmetry light) in the cockpit (other models are

the ATR 42-200 and -300 and the Cessna 500, 550, 553, 560, and 560XL airplanes).

Certification testing for these airplanes demonstrated that a cockpit light was not required

because a flap malfunction (including an asymmetry) did not constitute an unsafe flight

condition.

If the accident airplane had been equipped with a flap malfunction light in the cockpit,

the light would have illuminated amber in response to the flap asymmetry and would have

provided the flight crew with an immediate, definitive indication of a flap problem. The NTSB

concludes that, if the accident airplane had been equipped with a flap asymmetry light, as many

other ATR 42- and ATR 72-series airplanes are equipped, the illumination of that light would

likely have made the nature of the malfunction more salient to the flight crew and may have

triggered a more appropriate crew response. Therefore, the NTSB recommends that the FAA

require that all ATR 42-series airplanes be equipped with a flap asymmetry annunciator light if

they are not already so equipped.

2.6.3 Ice Evidence Probe

The accident airplane was not equipped with an IEP, which was an optional installation

for ATR airplanes that were not originally equipped with this device. Although an IEP provides a

surface upon which ice accumulation can be visually detected, the propeller spinners serve this

function for airplanes that are not equipped with an IEP, and the first officer indicated that she

observed ice on the spinners. Also, the AAS alerted the flight crew in a timely manner to apply

the procedures and checklists required when flying in icing conditions. Thus, the NTSB

concludes that the absence of an IEP, which was an optional installation, did not hinder the flight

crew’s ability to detect airframe ice accretion.

NTSB Aircraft Accident Report

2.7 Simulator Training for Icing Encounters

During the accident flight, both the flap asymmetry and the airframe ice accretion

adversely affected the airplane’s performance. Aerodynamic coefficient extraction indicated that

drag beyond the level expected for an uncontaminated airframe was present during cruise and the

final minutes of the flight. ATR estimated that, by the time that the flap asymmetry occurred, a

drag increase (due to ice) corresponding with about 23 percent of total power was present. The

EFS could not reproduce the accident airplane’s drag levels from ice accretion without

introducing unrealistic lift and pitch, roll, and yaw moments; therefore, the full extent of the

performance degradations could not be determined or duplicated in the EFS. No requirements

exist for simulators to be qualified with validated data, such as flight test and FDR data, for the

effects of airframe ice accumulation.

In this accident, the airplane’s stall warning system operated as designed and warned the

flight crew of the airplane’s proximity to the stall. However, other accidents and incidents have

occurred in which airframe ice accumulation has resulted in airplanes entering an aerodynamic

stall without any warning to the flight crews. For example, on January 2, 2006, a Saab SF340

operated by American Eagle Airlines, Inc., encountered icing conditions during the en-route

climb and departed controlled flight at an altitude of about 11,500 feet msl, losing about

5,000 feet of altitude before the pilots recovered control of the airplane.

148

The flight crew was

using the autopilot in vertical speed mode at the time that airplane control was lost, and

information from the FDR showed that the upset began before the stall warning activated.

The performance degradations associated with airframe ice accretion can result in an

aircraft slowing down much faster than normal and stalling at a much higher airspeed than

normal. Both conditions require pilot vigilance to recognize and react to the performance

degradations; however, both the American Eagle incident flight crew and this accident flight

crew failed to properly monitor and manage the airplanes’ airspeed. Although flight crews are

trained to use higher airspeeds during icing conditions, the effects of performance degradations

resulting from ice accretion are not realistically duplicated in a simulator.

The NTSB concludes that simulator-based training scenarios that realistically reflect

aircraft performance degradations that result from airframe ice accretion can better prepare flight

crews to effectively respond to decaying airspeed situations and other situations that can occur

during in-flight icing encounters. Therefore, the NTSB recommends that the FAA define and

codify minimum simulator model fidelity requirements for aerodynamic degradations resulting

from airframe ice accumulation. These requirements should be consistent with performance

degradations that the NTSB and other agencies have extracted during the investigations of icing

accidents and incidents.

The NTSB further recommends that the FAA, once the simulator model fidelity

requirements requested in Safety Recommendation A-11-46 are implemented, require that flight

crews of all aircraft certificated for flight in icing conditions be trained in flight training

simulators that meet these fidelity requirements. Such simulation training should emphasize the

148

The report for this incident, NTSB case number LAX06IA076, is available online at

<http://www.ntsb.gov/aviationquery/index.aspx>.

NTSB Aircraft Accident Report

following: (1) cues for recognizing changes in the aircraft’s flight characteristics as airframe

icing develops; (2) procedures for monitoring and maintaining appropriate airspeeds in icing

conditions, including the use of icing airspeed reference indices; and (3) procedures for

responding to decaying airspeed situations, stall protection system activation, and early stalls that

can occur without stall protection system activation.

NTSB Aircraft Accident Report

3. Conclusions

3.1 Findings

1. The captain and the first officer were certificated in accordance with Federal regulations and

were current and qualified in accordance with Empire Airlines’ training requirements. The

investigation found no evidence that the pilots’ performance was affected by any behavioral

or medical condition, or by the use of alcohol or drugs.

2. The airplane was loaded within weight and center of gravity limits and was maintained in

accordance with 14 Code of Federal Regulations Part 121.

3. The investigation found no evidence that indicated any mechanical anomaly with the

engines, and flight data recorder data indicated that the engines’ performance was consistent

with normal operations.

4. Performance data indicated that, when the flight crew commanded 15° flaps, a flap

asymmetry occurred with the left flaps extending 8° to 10° and the right flaps not extending.

The data further indicated that the flaps returned to a symmetric state (about 4.5°) about

25 seconds before ground impact.

5. Because of the extent of the impact and postaccident fire damage to the right flaps and flap

actuators, the reason for the airplane’s flap asymmetry could not be determined.

6. The airplane was controllable with the flap asymmetry and airframe ice contamination and

could have been maneuvered and landed safely if the appropriate airspeed had been

maintained.

7. The captain’s failure to immediately respond to the aural stall warning, the stick shaker, and

the terrain awareness and warning system warning resulted in his inability to arrest the

airplane’s descent and avoid impact with the ground.

8. The captain was adequately trained on how to respond to flap anomalies, and the captain’s

statement that the airplane had “no flaps” indicates that he had sufficient information to

recognize that he should immediately perform a go-around maneuver and apply the

appropriate procedure from the quick reference handbook that applied to all flap problems.

9. Although the captain indicated that he was concerned about the icing conditions, the presence

of airframe ice accretion within the capabilities of the airplane’s systems does not negate the

importance of adhering to standard operating procedures and performing a go-around

maneuver to respond to the multiple cues associated with an unstabilized approach including

excessive deviation from the glidepath, sink rate greater than 1,000 feet per minute, and

airspeed less than the required approach speed.

NTSB Aircraft Accident Report

10. Had the captain complied with standard operating procedures in response to the flap

anomaly, unstabilized approach, stick shaker, and terrain awareness and warning system

warning and initiated a go-around maneuver, the accident likely would not have occurred.

11. The first officer’s failure to maintain airspeed while acting as the pilot flying likely resulted

from being distracted by the flap anomaly, the captain’s actions in response to it, and the

control force inputs needed to maintain aircraft control.

12. The captain’s failure to call out the first officer’s airspeed deviations resulted directly from

his preoccupation with performing an inappropriate, nonstandard procedure in response to

the flap anomaly.

13. Although some of the airspeed bugs (including the internal bug) were not set to the

appropriate approach airspeeds and were not reset following recognition of the flap anomaly,

the flight crew had a sufficient reference to maintain the minimum safe airspeed because the

airspeed for a no-flap approach in icing conditions was correctly briefed as 143 knots, and

the red airspeed bugs were set near that value.

14. Reliance upon flight crew vigilance and stall warning systems may be inadequate to prevent

hazardous low-airspeed situations, and, had a low-airspeed alerting system been installed on

the airplane, it may have directed the flight crew’s attention to the decaying airspeed earlier

and provided an opportunity to take corrective action before the stall protection system

activated.

15. The first officer’s failure to assert herself to the captain and initiate a go-around maneuver

when she recognized the unstabilized approach likely resulted from the steep authority

gradient in the cockpit and the first officer’s minimal training on assertiveness; further, the

captain’s quick dismissal of the first officer’s go-around inquiry likely discouraged the first

officer from voicing her continued concerns and challenging the captain’s decision to

continue the unstabilized approach.

16. Role-playing exercises are essential for effective assertiveness training because such

exercises provide flight crews with opportunities for targeted practice of specific behaviors

and feedback that a lecture-based presentation format lacks.

17. The establishment of best practices for conducting both single and multiple emergency and

abnormal situations training needs to include training for the occurrence of these situations at

low altitudes because low-altitude scenarios require rapid, accurate assessment of abnormal

situations and appropriate prioritization of tasks.

18. Although the risk for fatigue existed at the time of the accident due to the window of

circadian low, the first officer took steps to mitigate the effects of fatigue, and her errors

during the flight can be explained by her lack of experience in both the airplane and in icing

conditions along with the distraction caused by the captain’s non-standard response to the

flap anomaly.

19. Due to the early morning hour of the accident flight and cumulative sleep debt, it is likely

that fatigue degraded the captain’s performance.

NTSB Aircraft Accident Report

20. Dispatching and operating an airplane in known icing conditions for which the airplane is not

certificated and has not demonstrated the ability to operate safely has the potential to reduce

or eliminate safety margins.

21. To most effectively ensure the safety of flight operations in icing conditions, pilots,

dispatchers, and flight followers must understand how the dangers of freezing drizzle and

freezing rain can affect their airplanes and must understand the differences between ground

deicing considerations and in-flight icing operations.

22. The presence of ice on the taxiways between the aircraft rescue and firefighting station and

the accident site minimally increased the response time; however, the emergency response

was timely and effective in suppressing the fire.

23. Timely information from the flight crew about the safety of the airplane occupants and the

presence of on-board hazardous materials cargo would improve the safety and efficiency of

the emergency response.

24. Because flight crews cannot always immediately communicate with air traffic control after

an accident, it is important that another method be developed to communicate information,

such as the number of occupants on board and the presence of hazardous materials, to aircraft

rescue and firefighting personnel upon initial notification of an accident.

25. An iced and inoperable mutual aid gate could extend the response time of mutual aid, which

could delay the delivery of medical attention to accident survivors and result in further fire

damage to property.

26. The aircraft performance monitoring system is a valuable low-airspeed alerting tool that can

enhance safety in all icing conditions.

27. If the accident aircraft had been equipped with a flap asymmetry light, as many other

ATR 42- and ATR 72-series airplanes are equipped, the illumination of that light would likely

have made the nature of the malfunction more salient to the flight crew and may have

triggered a more appropriate crew response.

28. The absence of an ice evidence probe, which was an optional installation, did not hinder the

flight crew’s ability to detect airframe ice accretion.

29. Simulator-based training scenarios that realistically reflect aircraft performance degradations

that result from airframe ice accretion can better prepare flight crews to effectively respond

to decaying airspeed situations and other situations that can occur during in-flight icing

encounters.

NTSB Aircraft Accident Report

3.2 Probable Cause

The National Transportation Safety Board determines that the probable cause of this

accident was the flight crew’s failure to monitor and maintain a minimum safe airspeed while

executing an instrument approach in icing conditions, which resulted in an aerodynamic stall at

low altitude. Contributing to the accident were 1) the flight crew’s failure to follow published

standard operating procedures in response to a flap anomaly, 2) the captain’s decision to

continue with the unstabilized approach, 3) the flight crew’s poor crew resource management,

and 4) fatigue due to the time of day in which the accident occurred and a cumulative sleep debt,

which likely impaired the captain’s performance.

NTSB Aircraft Accident Report

4. Recommendations

As a result of this investigation, the National Transportation Safety Board makes the

following recommendations to the Federal Aviation Administration:

Require that role-playing or simulator-based exercises that teach first officers to

assertively voice their concerns and that teach captains to develop a leadership

style that supports first officer assertiveness be included as part of the already

required crew resource management training for 14 Code of Federal Regulations

Part 121, 135, and 91 subpart K pilots. (A-11-39)

Prohibit all 14 Code of Federal Regulations Part 121, 135, and 91 subpart K

operators of pneumatic deice boot-equipped airplanes from dispatching or

deliberately operating these airplanes in known freezing rain or freezing drizzle of

any intensity, unless the airplane manufacturer has demonstrated that the airplane

model can safely operate in those conditions. (A-11-40)

Review the approved pilot, dispatcher, and flight follower training programs and

procedures for all 14 Code of Federal Regulations Part 121, 135, and 91

subpart K operators and require revisions to the programs and procedures, as

necessary, to include standardized training and aircraft-specific information to

educate pilots, dispatchers, and flight followers of the dangers of flight operations

in freezing precipitation and of the differences between ground deicing

considerations and in-flight icing operations. (A-11-41)

Develop a method to quickly communicate information regarding the number of

persons on board and the presence of hazardous materials to emergency

responders when airport emergency response or search and rescue is activated.

(A-11-42)

Amend Advisory Circular 150/5200-30C to include guidance on monitoring and

ensuring the operability of emergency response and mutual aid gates during

winter operations. (A-11-43)

Require all operators of Avions de Transport Régional Aerospatiale Alenia

ATR 42- and ATR 72-series airplanes to retrofit the airplanes with an aircraft

performance monitoring system if they are not already so equipped. (A-11-44)

Require all Avions de Transport Régional Aerospatiale Alenia ATR 42-series

airplanes to be equipped with a flap asymmetry annunciator light if they are not

already so equipped. (A-11-45)

Define and codify minimum simulator model fidelity requirements for

aerodynamic degradations resulting from airframe ice accumulation. These

requirements should be consistent with performance degradations that the

NTSB Aircraft Accident Report

National Transportation Safety Board and other agencies have extracted during

the investigations of icing accidents and incidents. (A-11-46)

Once the simulator model fidelity requirements requested in Safety

Recommendation A-11-46 are implemented, require that flight crews of all

aircraft certificated for flight in icing conditions be trained in flight training

simulators that meet these fidelity requirements. Such simulation training should

emphasize the following: (1) cues for recognizing changes in the aircraft’s flight

characteristics as airframe icing develops; (2) procedures for monitoring and

maintaining appropriate airspeeds in icing conditions, including the use of icing

airspeed reference indices; and (3) procedures for responding to decaying airspeed

situations, stall protection system activation, and early stalls that can occur

without stall protection system activation. (A-11-47)

BY THE NATIONAL TRANSPORTATION SAFETY BOARD

DEBORAH A.P. HERSMAN ROBERT L. SUMWALT

Chairman Member

CHRISTOPHER A. HART MARK R. ROSEKIND

Vice Chairman Member

EARL F. WEENER

Member

Adopted: April 26, 2011

Vice Chairman Hart and Members Rosekind and Weener filed the following statements

on May 9, 2011.

NTSB Aircraft Accident Report

Board Member Statements

Vice Chairman Christopher A. Hart, dissenting:

I dissent from this report because (a) it is internally inconsistent, and (b) it concludes,

without adequate basis, that fatigue contributed to the accident.

Internal Inconsistency. The report is internally inconsistent because, on one hand, it

emphasizes the general rule, and the probable cause asserts, that the captain should have

discontinued his approach when it became unstable. On the other hand, it recommends that the

FAA should prohibit these types of airplanes “from dispatching or deliberately operating” in

precisely the type of icing conditions that it says this captain should have re-entered in his

go-around. Either the conditions were flyable and should have been re-entered on a go-around, or

the conditions were not flyable and the captain appropriately continued his approach, despite

being unstable, but not both.

I believe that the evidence supports the conclusion that the

conditions were flyable and the captain should have gone around.

More specifically, the evidence indicates that the icing conditions encountered during the

approach were well within the capability of the airplane and the crew. The crew had

appropriately engaged the necessary ice protection systems, they never expressed any concern

about icing, and the airplane had plenty of performance despite the icing. Thus, instead of being

an icing accident, this was an accident involving an unstable approach that resulted from

untimely and excessive corrections, along with poor CRM, during an instrument approach, and

there is no basis in the report for concluding that icing was in any way causal or even

contributory to this accident.

With a few small exceptions, the flight crew appeared to be working well together and

was generally “ahead of the curve” until they began to descend when the autopilot intercepted

the glide slope. Because the copilot (who was the pilot flying) did not reduce the power as their

descent began, the airplane began to accelerate. Her response to the acceleration was to reduce

power very substantially, essentially to flight idle. Making large corrections during an instrument

approach is not good practice because large corrections usually necessitate subsequent large

corrections in the opposite direction, which, as occurred here, can easily lead to an unstable

approach. About fifteen seconds after the autopilot intercepted the glide slope and they began

descending, and about ten seconds after she substantially reduced the power, something

happened – the investigation apparently was not able to determine with certainty what – that

caused the copilot to ask, at 0435:03, “what the heck is going on?” Her question may have been

prompted by the airplane’s reaction(s) to their effort to extend the flaps because, one second

later, the captain responded, “You know what? We have no flaps.”

Consequently, as they were rapidly decelerating during their descent along the glide

slope, her inquiry and the captain’s response about the flaps caused both pilots to divert their

The recommended prohibition against “dispatching” into icing conditions is proactive and appropriate, and

does not create an inconsistency. The recommended prohibition against “deliberately operating” in icing conditions,

however, flies in the face of asserting that the captain should have aborted his approach and re-entered the prohibited

icing conditions.

NTSB Aircraft Accident Report

attention to the flap problem. The captain, as pilot monitoring, appropriately began to trouble

shoot, but inappropriately began looking at circuit breakers rather than the flap malfunction

checklist. The copilot, however, was also distracted by the flap problem, which diverted her

attention away from their decreasing airspeed until it decreased enough to activate the stick

shaker and disconnect the autopilot. The copilot’s diversion of her attention was contrary to good

CRM practice that in the event of a malfunction, one pilot should explore the malfunction,

preferably starting with the checklist where appropriate, while the other pilot flies the airplane.

After the autopilot disconnected, the approach became unstable, they did not maintain adequate

airspeed, even after the stick shaker activated, and the airplane stalled and crashed.

Hence, this accident actually began with the copilot’s failure to reduce power when they

intercepted the glide slope, followed by her excessive power reduction in response to the

resulting acceleration, followed by the distraction caused by the flap problem that diverted her

attention – until the first stick shaker event – from the need to restore adequate power to prevent

further loss of speed. Because none of the events in this causal chain related to or were caused by

icing, this is not an icing accident, and there is no basis from this accident for any

recommendations regarding icing.

The higher stall speed due to icing does not undercut my conclusion that this is not an

icing accident. The copilot’s substantial power reduction caused an initial deceleration of about

2 knots per second, and I am not convinced, for two reasons, that the few additional seconds it

would have taken for them to decelerate to a lower (non-iced) stall speed might have produced a

different result. First, I see no reason to believe that those extra few seconds would have been

enough to cause them to return their attention – before the stall – to the need to arrest their rapid

speed decrease. Second, given their lack of appropriate response – not one, not two, but three

times – to the stick shaker, I see no basis for concluding that that those extra seconds would have

altered their non-response to the stick shaker.

Consequently, I do not believe that this was an icing accident, and

‐ The fact that they were in icing conditions is not relevant to, and should not be in, the

probable cause;

‐ The recommendations relating to in-flight icing, irrespective of whether they are

good ideas, are not supported by this accident; and

‐ Because Recommendation No. 2 asks the FAA to prohibit operation in the types of

icing conditions that were present in this accident, the probable cause should not

assert that the captain should have discontinued his approach and ascended back into

those icing conditions.

Fatigue. The probable cause concludes that fatigue “likely impaired the captain’s

performance.” Even if the crew had been fatigued, which they probably were to some extent, I

do not see any basis in the report for concluding that fatigue resulted in impairment sufficient to

cause or contribute to this accident.

There are several aspects about fatigue that necessitate extreme caution before including

it as part of the probable cause. To name a few, fatigue:

NTSB Aircraft Accident Report

‐ Is not either “on” or “off,” but can range from none, to extreme, to anywhere in

between;

‐ Can be quite variant from day to day, and can be present one day, enough to cause

problems, but not be present to that extent the next day; and it can even be quite

variant in the course of a single flight, depending upon the duration of the flight and

events during the flight;

‐ Is very widespread – at any point in time, a large percentage of the population is

probably suffering from fatigue to some extent;

‐ Is multidimensional – it is usually referred to in relation to sleep deficit, but it is also

affected by diet, exercise, smoking, alcohol use, age, and other factors, and the

impacts and interrelationships among and between those factors are often not well

understood;

‐ Is not directly measurable – it can be ascertained only indirectly, on the basis of

actions and behavior;

‐ Has effects that can arguably be counteracted by processes that are designed to reduce

the need to analyze and think, such as standard operating procedures and checklists;

‐ May be significantly mitigated, at least temporarily to some extent, by an adrenalin

rush in a life-threatening situation; and

‐ May be overwhelmed by the effects of a life-threatening situation.

Given all of these characteristics and uncertainties, caution is warranted regarding the

inclusion of fatigue as part of probable cause because, among other reasons, the incomplete

understanding of, and inability to measure, fatigue and its effects make it very difficult to

conclude that fatigue is causative; and given how widespread fatigue is, query how much our

inclusion of it in the probable cause can help us achieve our ultimate objective, which is the

issuance of safety-improving recommendations.

In this accident, there is probably ample basis for concluding, based upon sleep alone,

that both crewmembers were fatigued to some extent. Less ascertainable, however, is whether

their performance was degraded by fatigue enough to be causal or contributory to the accident.

As noted above, the flight crew appeared to be working well together and was generally “ahead

of the curve,” with a few small exceptions, until the autopilot intercepted the glide slope and they

neglected to reduce power when they began their descent. Their satisfactory behavior during

most of the flight could support the conclusion that their fatigue was not enough to impair their

performance, at least in the earlier stages of the flight. Another possible conclusion, however, is

that even if their fatigue were sufficient to impair their performance, their use of standard

operating procedures and checklists earlier in the flight may have masked the impairment effects.

Yet a third possible conclusion is that, even if fatigue did not measurably impair their

performance earlier in the flight, it degraded their performance later in the flight enough to be

causal or contributory. These varying and not necessarily consistent conclusions that can be

drawn from the evidence demonstrate the difficulty of labeling fatigue as causal to performance

impairment.

NTSB Aircraft Accident Report

Another complication derives from the fact that this was a life-threatening experience for

the crew. We can probably conclude with some confidence that the crew in this accident

considered their situation to be life-threatening, perhaps starting with her query, “What the heck

is going on?” at 0435:03, followed by his response, at 0435:04, “You know what? We have no

flaps.” Her question and his answer occurred only 84 seconds before the airplane impacted the

ground, during about half of which time they were out of the clouds and were able to see the

lighted runway and ground rapidly approaching.

At least two possibilities come to mind from the fact that the experience was life-

threatening. First, the crew probably experienced an adrenaline rush, and the adrenaline may

have reduced the impairing effects of fatigue, at least temporarily, in the critical final seconds

before impact. I am not aware of any literature on this issue, but intuitively it seems very

possible that adrenaline could, at least temporarily, counteract some of the potentially impairing

effects of fatigue. Second, their behavior in the final seconds before impact may have been so

overwhelmingly influenced by their knowledge that they were in a life-threatening situation that

the effects of fatigue were insignificant by comparison.

Given this extent of uncertainty, I do not believe that we have adequate basis for

concluding in the probable cause that fatigue “likely impaired the captain’s performance.”

NTSB Aircraft Accident Report

Member Mark R. Rosekind, concurring:

This accident has generated a constructive debate among the staff and Board Members on

several key issues, one of which is fatigue. Such debates support a thorough analysis and are at

the core of the NTSB’s mission. I commend the staff and fellow Board Members for raising

these issues.

This accident involved night shift work, circadian lows, and misconceptions regarding

the ability to adjust to overnight operations. As the report adequately sets forth, this accident

occurred during the window of circadian low at about 0437. Well-established scientific research

demonstrates that this time is associated with severe performance reductions, and increased

errors, incidents, and accidents. As a result, accidents involving pilot error that occur within this

window of time should be thoroughly examined for fatigue. While mitigation strategies can be

employed to somewhat counter the performance impairments caused by fatigue, overnight

operations will always be inherently more at risk due to the human body’s internal circadian

clock.

Scientific research has shown that the circadian clock does not fully adjust to night

operations over time without very specific interventions that require ideal conditions and are

very difficult to achieve. These targeted interventions require specific manipulation of light/dark

exposure and sleep which runs counter to our society.

In this accident, both pilots altered their sleep schedules in an effort to adjust their

circadian rhythms to their night flying operations. There was no information that either pilot had

received education and training about these interventions or had knowledge of the specific

actions required to shift circadian rhythms. The First Officer made a more extensive effort to

adjust her schedule and maintained it for a longer period of time. Despite these actions, no

information indicated she sufficiently employed mitigation strategies to effectively adjust her

circadian clock. For example, there is insufficient information to assess her light exposure or

whether she slept continuously during her sleep opportunity periods.

While I voted for and supported the conclusions set forth in the report, this concurring

statement is written, in part, to disagree with the conclusion that the First Officer’s errors “can be

explained by her lack of experience in both the airplane and in icing conditions along with the

distraction caused by the captain’s nonstandard response to the flap anomaly” (Conclusion 18).

The evidence that fatigue contributed to her errors is no less than the evidence suggesting that

her lack of experience or the Captain’s nonstandard response contributed to the accident.

Finally, this accident exemplifies the increased safety-risks associated with overnight

shifts and operations during the window of circadian low (at a minimum between 0300 and

0500) and should guide hours of service regulations, scheduling policies and practices, and

safety –sensitive operations affected by circadian factors.

NTSB Aircraft Accident Report

Member Earl F. Weener, concurring in part and dissenting in part:

Generally, I agree with the report and recommendations of Notation 8093B, as presented

to the Board on Tuesday, April 26, 2011. However, I cannot support the Probable Cause

statement, as adopted by the Board, or conclusion number 19.

As discussed at the Board meeting, I am unconvinced the investigation or final report

provides sufficient evidence to conclude that fatigue impaired the captain’s performance to the

extent it can be considered a cause of this accident, not even as a contributory factor.

I agree the captain may very well have experienced fatigue, and fatigue may have even

played a role in his performance. However, based on the evidence provided in this case, and as

presented in the final report, the causal connection between fatigue and the actions leading to this

accident has not been made.

Probable, as the word implies, means more than simply a possibility. The report initially

states that sleep needs vary by individual (p. 66). It then identifies several possibilities to explain

the captain’s performance, including past behavior and stress, along with fatigue. From this

observation the report concludes that the captain was “likely experiencing some fatigue at the

time of the accident,” (emphasis added) (p. 67), and that “fatigue likely degraded his

performance” (emphasis added) (p. 68). However, the report fails to provide an explanation for

why the possibility of fatigue is any more likely or probable than the other possibilities initially

identified. There is an unaccounted leap from several possibilities to explain the captain’s

performance to the singular conclusion that fatigue likely degraded the captain’s performance to

the extent it became a contributing causal factor of the accident. Further, the report fails to

explain how “likely experiencing some fatigue” becomes “likely degraded his performance.”

These conclusions, possibilities at best, do not support a finding of probability.

The immediate sequence of events leading to the accident began when the First Officer

(FO) mismanaged thrust at the interception of the glide slope. This was followed by loss of

airspeed leading to activation of stick-shaker and the subsequent disconnect of the autopilot. The

FO then had difficulty tracking the ILS localizer and glideslope and quickly deviated from the

ILS approach path to the point where the Captain asked if he should take over the approach. At

that point, the captain took manual control of an airplane that was likely out of trim and at a

power setting inconsistent with the intended ILS flight path. The airmanship errors committed by

the FO were certainly on the order of those subsequently made by the captain. Yet, staff

conclude that only the captain’s errors were likely caused by fatigue, while the FO’s were not.

In my opinion, nothing in the report substantiates this difference.

The report does discuss, in great detail, the FO’s actions, which, in turn, substantiates the

conclusion that her errors can be explained by lack of experience. However, the same level of

detail and analysis is lacking in terms of evaluating the captain’s actions. In fact, it is quite a

stark comparison. Also, the report cites a number of authoritative studies concerning the effects

of fatigue on human performance. However, the correlation of the studies to the crew’s

performance, is ambiguous. The report even acknowledges that the flight crew’s failure to

monitor air speed, the identified probable cause of this accident, could be consistent with the

known effects of fatigue; and at the same time, the report notes, the same error has been made by

NTSB Aircraft Accident Report

other crew who were not fatigued (p. 68). Again, at best, fatigue is one of several possibilities

that could explain the captain’s performance.

Further, when questioned on these points during the Board Meeting, staff again stated

that fatigue was one of a number of possibilities that could explain the captain’s performance,

and that indeed, it could also be explained by past behavior. The critical basis staff provided for

concluding that fatigue was a likely cause of the behavior, was an analysis derived by using the

captain’s own judgment that he required 6-8 hours of sleep to “feel rested,” and choosing to use

the more conservative end of this range, 8 hours, to arrive at a calculation of a sleep debt of 4

hours. However, I point out that in the initial draft of this report, fatigue was not identified as a

causal issue in this investigation. In fact, in briefings with staff in preparation for the Board

meeting, staff made a point of stating that they did analyze the role of fatigue in this accident and

ruled it out as a factor, as demonstrated in the initial report. When questioned at the Board

meeting on why this determination changed, staff explained that in the initial analysis they chose

to use the lower end of the captain’s sleep need range, 6 hours, to conduct the analysis and

subsequently determined fatigue did not play a role. Candidly, when the outcome of whether a

factor contributed to the cause of an accident is dependent on choosing from within a range, the

analysis is inconclusive.

Finally, I note the probable cause statement in the final draft of the report states the

probable cause for this accident was the flight crew’s failure to monitor and maintain a minimum

safe airspeed while executing an instrument approach in icing conditions. I further note the

original contributing factors to the accident in the initial draft report were limited to: 1) the flight

crew’s failure to follow published standard operating procedures in response to a flap anomaly;

2) the captain’s decision to continue with the unstabilized approach; and 3) the flight crew’s poor

crew resource management. All of these factors are supported by observable actions or inactions.

They were objectively derived, and not one of them is described as simply “likely” to have

occurred.

In sum, the report and probable cause, as initially proposed, provided a rational and

substantiated explanation for the cause of the accident. Although fatigue may have played a role

in the captain’s performance during the accident sequence, the final report does not sufficiently

make the case that fatigue played a causal role in the event.

NTSB Aircraft Accident Report

5. Appendixes

Appendix A

Investigation and Hearing

Investigation

The National Transportation Safety Board (NTSB) was notified about the accident on

January 27, 2009. Staff from the NTSB arrived on scene that same day and remained there until

January 30, 2009, to conduct the field portion of the investigation. No Board Member

accompanied the team.

In accordance with the provisions of Annex 13 to the Convention on International Civil

Aviation, the NTSB’s counterpart agencies in France and Canada, the Bureau d’Enquêtes et

d’Analyses (BEA) and the Transportation Safety Board (TSB), participated in the investigation

as the representatives of the State of Design and Manufacture (Airframe and Powerplants,

respectively). Avions de Transport Régional (ATR) and the European Aviation Safety Agency

(EASA) participated as technical advisors to the BEA. Pratt & Whitney, Canada, participated as

a technical advisor to the TSB.

Parties to the investigation were the Federal Aviation Administration (FAA); Empire

Airlines, Inc.; and FedEx Corporation.

Public Hearing

A public hearing was held on September 22 and 23, 2009, in Washington, D.C. Chairman

Deborah A.P. Hersman presided over the hearing.

The issues presented at the public hearing included flight crew procedures, crew resource

management and decision making, the airline’s training program for flight in icing conditions,

and airplane design modifications to monitor and mitigate the effects of icing on airplane

performance. Parties to the public hearing were the FAA, Empire Airlines, and ATR.

NTSB Aircraft Accident Report

Appendix B

Cockpit Voice Recorder Transcript

The following is a transcript of transcript of a Fairchild Model A-100A tape CVR, serial

number 59653, installed on an Empire Airlines ATR-42-320 (N902FX), which crashed during

landing at Preston Smith International Airport in Lubbock, Texas.

LEGEND

CAM Cockpit area microphone voice or sound source

HOT Flight crew audio panel voice or sound source

RDO Radio transmissions from N902FX

CTR Radio transmission from Dallas center controller

APR Radio transmission from the Lubbock approach controller

TWR Radio transmission from the Lubbock tower controller

OPS Radio transmission from Lubbock FedEx Operations

TAWS Terrain Awareness and Warning System sound source

-1 Voice identified as the captain

-2 Voice identified as the first officer

-? Voice unidentified

* Unintelligible word

# Expletive

@ Non-pertinent word

( ) Questionable insertion

[ ] Editorial insertion

Note 1: Times are expressed in CST.

Note 2: Generally, only radio transmissions to and from the accident aircraft were transcribed.

Note 3: Words shown with excess vowels, letters, or drawn out syllables are a phonetic representation of the words

as spoken.

Note 4: A non-pertinent word, where noted, refers to a word not directly related to the operation, control or condition

of the aircraft.

NTSB Aircraft Accident Report

CVR Quality Rating Scale

The levels of recording quality are characterized by the following traits of the cockpit voice recorder

information:

Excellent Quality Virtually all of the crew conversations could be accurately and easily understood.

The transcript that was developed may indicate only one or two words that were

not intelligible. Any loss in the transcript is usually attributed to simultaneous

cockpit/radio transmissions that obscure each other.

Good Quality Most of the crew conversations could be accurately and easily understood. The

transcript that was developed may indicate several words or phrases that were

not intelligible. Any loss in the transcript can be attributed to minor technical

deficiencies or momentary dropouts in the recording system or to a large number

of simultaneous cockpit/radio transmissions that obscure each other.

Fair Quality The majority of the crew conversations were intelligible. The transcript that was

developed may indicate passages where conversations were unintelligible or

fragmented. This type of recording is usually caused by cockpit noise that

obscures portions of the voice signals or by a minor electrical or mechanical

failure of the CVR system that distorts or obscures the audio information.

Poor Quality Extraordinary means had to be used to make some of the crew conversations

intelligible. The transcript that was developed may indicate fragmented phrases

and conversations and may indicate extensive passages where conversations

were missing or unintelligible. This type of recording is usually caused by a

combination of a high cockpit noise level with a low voice signal (poor signal-to-

noise ratio) or by a mechanical or electrical failure of the CVR system that

severely distorts or obscures the audio information.

Unusable Crew conversations may be discerned, but neither ordinary nor extraordinary

means made it possible to develop a meaningful transcript of the conversations.

This type of recording is usually caused by an almost total mechanical or

electrical failure of the CVR system.

INTRA-AIRCRAFT COMMUNICATION AIR-GROUND COMMUNICATION

TIME and TIME and

SOURCE CONTENT SOURCE CONTENT