Eastern Air Lines Runway Excursion at La Guardia, final report

Several failures in close succession by a jetliner’s flight crew were the probable cause of Oct. 27, 2016, runway excursion at La Guardia Airport, according to the National Transportation Safety Board’s final report issued September 21, 2017. Some recurrent training issues were identified as contributing factors.


Photo (C) ABC news


History of Flight

Landing-flare/touchdown:  Landing area overshoot

Landing-landing roll:  Runway excursion (Defining event)

On October 27, 2016, about 1942 eastern daylight time, Eastern Air Lines flight 3452, a Boeing 737-700, N923CL, overran runway 22 during the landing roll at LaGuardia Airport (KLGA), Flushing, Queens, New York. The airplane travelled through the right forward corner of the engineered materials arresting system (EMAS) at the departure end of the runway and came to rest off the right side of the EMAS. The 2 certificated airline transport pilots, 7 cabin crewmembers, and 39 passengers were not injured and evacuated the airplane via airstairs.

The airplane sustained minor damage. The charter flight was operating under the provisions of 14 Code of Federal Regulations Part 121. Night instrument flight rules conditions prevailed at the airport at the time of the incident, and an instrument flight rules flight plan was filed for the flight, which originated at Fort Dodge Regional Airport (KFOD), Fort Dodge, Iowa, about 1623 central daylight time.

The first leg of the trip began on October 14, 2016, and the captain and first officer were paired from then to the incident. In post-incident statements, the flight crew indicated that the captain was the pilot monitoring (PM) for the incident flight, and the first officer was the pilot flying (PF). The first officer reported that the autopilot and autothrottles were engaged beginning about 2,500 ft after their takeoff from KFOD. Both pilots stated that the en route portion of the flight and the descent into the terminal area were uneventful but they encountered moderate to heavy rain during the final 15 minutes of the flight.

According to information from the airplane’s cockpit voice recorder (CVR), the first officer partially briefed the instrument landing system (ILS) approach for runway 13 beginning about 1848, indicating an autobrake setting of 3 and a 30º flap setting. ATIS information “Bravo” was current at that time and indicated visibility 3 miles in rain, ceiling 1,500 ft broken, overcast at 2,200 ft, wind from 130º at 9 knots, and that braking action advisories were in effect. About 1852, the first officer began briefing the ILS approach for runway 22 after the captain clarified, based on the ATIS recording, that runway 13 was being used for departures.

About 1902, as the airplane descended through 18,000 ft msl, the flight crew completed the approach briefing for runway 22, with the same autobrake and flap setting as indicated earlier, as well as the decision altitude and visibility required for the approach, the touchdown zone elevation, and a reference speed (Vref) of 137 knots. ATIS information “Charlie” was current at that time and indicated visibility 3 miles in rain, ceiling 900 ft broken, overcast at 1,500 ft, and wind from 120º at 9 knots.

The flight crew also discussed the captain manually deploying the speed brakes (the airplane’s automatic speed brake module had been deactivated 2 days before the incident and deferred in accordance with the company’s minimum equipment list (MEL), with corrective action scheduled for November 4, 2016). In reference to the manual deployment of the speed brakes, the captain stated at 1902:44.5 “you’re gonna do these. I’m gonna do this” to which the first officer replied “[that] is correct.”

About 1927, the flight was provided vectors to the final approach course for the ILS approach to runway 22. About 1936, the flight was cleared for the approach. The first officer then called for the landing gear to be extended and the flaps set at 15º. About 1937, the captain stated that the localizer and glideslope were captured. About 1938, as the airplane neared the final approach fix, the flight crew completed the landing checklist and configured the airplane for landing, with flaps set to 30º.

The CVR indicates that the captain pointed out the approach lights about 1939. The first officer reported, and flight data recorder (FDR) data indicate, that about 1940:12, he disconnected the autopilot when the airplane’s altitude was about 300 ft radio altitude, as required by Eastern Air Lines standard operating procedure. FDR data indicate that the first officer disconnected the autothrottles about 1940:19.

FDR data indicate that, shortly after the first officer disconnected the autopilot and autothrottles (about 300 ft radio altitude), the airplane began to increasingly deviate above the glideslope beam and crossed the threshold at a height consistent with the threshold crossing height of the VGSI, which was not coincident with the glide slope beam. CVR data indicate that between 1940:35 and 1940:46, the enhanced ground proximity warning system alerted the decreasing altitude in increments of 10, beginning at 50 ft. The pitch attitude started to increase in the flare from 2.8° at a radio altitude of about 38 ft. After the 20-ft alert, the captain stated “down” at 1940:43.3. After the 10-ft alert, the captain stated, “down down down down you’re three thousand feet remaining” at 1940:46.6. There was no callout of spoilers or thrust reversers during the rollout on the CVR.

FDR data and performance calculations indicate that the airplane crossed the runway threshold at a radio altitude of 66 ft, with an increasing glideslope deviation and a descent rate of about 750 ft per minute. When the airplane had travelled about 2,500 ft beyond the runway threshold, its descent rate decreased to near zero, and it floated before touching down. The captain later reported that the descent to the touchdown zone was normal until the flare. He stated that the airplane floated initially in the flare, which prompted the captain to tell the first officer to “get it down.”

The first officer recalled hearing the captain’s instruction to “put [the airplane] down” during the flare but was not certain how far down the runway the airplane touched down. FDR data indicate that at 1940:51.8, the airplane’s main landing gear touched down; maximum manual wheel brakes were applied at main gear touchdown. The throttles were not fully reduced to idle until about 16 seconds after the flare was initiated, and after the airplane had touched down.

The touchdown point was about 4,242 ft beyond the threshold of the 7,001-ft-long runway. The nose gear initially touched down about 2 seconds after the main landing gear but rebounded into the air due to aft control column input. The nose gear touched down a second and final time at 1940:56.8.

The captain reported that, as briefed, he manually deployed the speed brakes, which FDR data indicate were manually extended to full at 1940:56.3, about 4.5 seconds after the main landing gear touched down and the airplane had travelled about 1,250 ft farther down the runway from the touchdown point. At 1940:59.8, when the airplane had travelled about 1,650 ft down the runway from the touchdown point (and 5,892 ft from the threshold), maximum reverse thrust was commanded. The captain reported that he saw the end of the runway approaching and began to apply maximum braking, as well as right rudder because he thought it would be better to veer to the right rather than continue straight to the road beyond the end of the runway.

The first officer reported that the captain did not, as required in the operator’s procedures, tell him that he was attempting to brake and steer the airplane during the landing rollout, and no such callout is recorded on the CVR. The first officer stated that the airplane was pulling to the right “really hard,” which prompted him to apply left rudder. He reported that the left rudder input was counter to his expectation due to a 9-knot crosswind from the left, which he expected to counteract with right rudder input. He attempted to maintain alignment with the runway centerline by applying left rudder and overriding the autobrakes with pressure on the brake pedal.

At 1941:08.3, the CVR recorded the sound of rumbling, consistent with the airplane exiting the runway. The airplane then entered the EMAS about 35 knots groundspeed and came to rest 172 ft beyond the end of the runway and to the right of the EMAS. Review of the CVR recording revealed that, after the airplane came to a stop, the first officer twice remarked that they should have conducted a go-around, and the captain agreed. The first officer later reported that he did not believe the approach or landing were abnormal at the time. The captain later stated that he should have called for a go-around when the airplane floated during the flare.


Photo (C) REUTERS Lucas Jackson

Flight Crew Information

Pilot 1

Pilot 2

Meteorological Information and Flight Plan


At 1851 EDT, (ASOS) at KLGA reported the wind from 090° true at 9 knots, visibility of 3 statute miles (sm), moderate rain, ceiling broken at 900 ft agl, overcast clouds at 1,500 ft agl, temperature of 13°C and a dew point temperature of 11°C, and altimeter setting of 30.14 inches of mercury. Remarks included: surface visibility of 4 sm, precipitation accumulation of 0.14 inch since 1751 EDT.

At 1951 EDT, KLGA ASOS reported the wind from 100° true at 10 knots with gusts to 15 knots, visibility of 3 sm, moderate rain, mist, ceiling overcast at 1,000 ft agl, temperature of 13°C and a dew point temperature of 12°C, and an altimeter setting of 30.10 inches of mercury. Remarks included: surface visibility of 4 sm, precipitation accumulation of 0.32 inch since 1851 EDT, precipitation accumulation of 0.61 inch during previous 3 hours.

Wreckage and Impact Information

Crew:  11 Injuries: None

Passenger: 37 Injuries: None

Ground Injuries: N/A

Total 48 Injuries: None

Aircraft Damage: Minor

Aircraft Fire: None

Aircraft Explosion: None

Latitude, Longitude: 40.769167, -73.885000

As a result of the airplane’s travel through the EMAS, pulverized EMAS material (a gray, powdery residue) was noted on portions of the airplane’s exterior during postincident examination. The lower and forward portions of the airplane—fuselage, landing gear, and antennas—were coated with a dried residue resulting from the mixture of the EMAS material and rainwater. In addition, pieces of a matting material used in the EMAS were found in various locations on the airplane.

No damage or anomalies were noted during the visual examination of the nosewheel landing gear and associated assemblies. A preliminary visual examination of the main landing gear strut, doors, assemblies, associated hydraulic lines, and antiskid components did not reveal evidence of physical damage. However, after the airplane was cleaned of EMAS debris and the main landing gears were retracted, damage was noted on the underside of each gear strut. The operator indicated that the lower wire bundle support brackets for the left and right main landing gear were both damaged, as well as the wire conduit sleeve on the left main landing gear.

Each of the four main wheel tires showed cut damage in addition to normal wear. None of the observed cuts were deep enough to reach the tire treads. No flat spots or other evidence of hydroplaning was noted on any of the tires. Examination of the four brake assemblies found no evidence of damage or hydraulic leaks. No evidence of a hydraulic power malfunction or damage to any of the visible hydraulic lines was noted.

Both engines showed evidence of EMAS material and matting on the engine inlet and internal components. The No. 1 engine sustained fan blade damage, including four blades bent in the direction opposite of rotation, at the tip corner. No visible blade damage was noted on the No. 2 engine. Visual examination of the thrust reversers found no preincident anomalies. The operator later reported that, after cleaning and deploying the thrust reversers, damage was found on the inboard thrust reverser sleeves and blocker doors for both engines.


Photo (C) NTSB

Examination of the speed brake control components on the incident airplane noted the speed brake handle positioned full forward. All spoiler panels, including the ground spoilers, were found in the down or retracted position. No damage was noted to any of the ground spoilers.


No problems with communications equipment were reported.

Flight Recorders

The airplane was equipped with a cockpit voice recorder (CVR) and a flight data recorder (FDR). Both recorders were removed from the airplane and retained by the NTSB for further examination and readout at the NTSB’s Recorder Laboratory in Washington, DC. The recorders showed no signs of damage.

Cockpit Voice Recorder

The CVR, a Honeywell 6022, serial number 3452, was a solid-state CVR that recorded 120 minutes of digital audio. It was played back normally without difficulty and contained excellent quality audio information. The recording was transcribed in two parts focusing on the en route approach briefing and the approach, landing, and events thereafter until the end of the recording. Part one began at 18:48:06 EDT, when flight 3452 was en route at FL390, and continued until 1902:52 EDT. Part two began at 1918:01 EDT and ended at 1948:32 EDT (the end of the recording

Flight Data Recorder

The FDR, a Honeywell 4700, serial number SSFDR-16936, recorded airplane flight information in digital format using solid-state flash memory as the recording medium. The FDR could record a minimum of 25 hours of flight data and was configured to record 256 12-bit words of digital information every second. The FDR was designed to meet the crash survivability requirements of Technical Standard Order C-124.

Data from the FDR were extracted normally. The event flight was the last flight of the recording, and its duration was about 2 hours and 19 minutes.

Medical And Pathological Information

Eastern Air Lines conducted drug and alcohol testing for both pilots about 6 hours after the incident. Test results were negative for alcohol and major drugs of abuse.

Organizational And Management Information

Company Overview and Management Organization

Eastern Air Lines, Inc., received certification to operate as a Part 121 supplemental carrier on May 15, 2015. Subsequently, Eastern Air Lines began scheduled charter services to Havana and four other cities in Cuba. Before the incident, the airline also launched charter service to other Latin American and Caribbean destinations. The airline’s sole base of operations was at Miami International Airport, Miami, Florida, at the time of the incident. It employed 64 pilots and had a fleet of five Boeing 737 airplanes, including the incident airplane; the other four airplanes were Boeing 737-800 series.

The airline’s vice president of flight operations was responsible for the flying operations of the airline, flight crew training, the operations control center (OCC), and ground operations. The chief pilot, manager of flight operations training, director of inflight, OCC director, manager of flight standards, and manager of charter operations all reported to the vice president of flight operations.

At the time of the incident, Eastern Air Lines’ director of safety and security reported directly to the chief executive officer and was the only staffed position in the safety department. The director of safety and security had been hired about 2 weeks before the incident and was in the process of being trained by his predecessor, who had held the position from 2013 until September 2016. While he was being trained, the vice president of regulatory compliance served as the acting director of safety and security.

According to the vice president of flight operations and the manager of flight operations training, the Boeing 737 Flight Crew Training Manual and the Boeing 737 Flight Crew Operations Manual were used as the airline’s systems training material and procedures manual, respectively.

Safety Management

The FAA approved Eastern Air Lines’ safety management system (SMS) implementation plan in February 2016. The first segment of implementation included administering the SMS implementation plan and developing a tool (Aviation Resource Management Solutions) that was designed to help the company with safety risk assessment, assurance, and risk management. The former director of safety and security stated that, at the time of the incident, the first segment of the implementation was not fully realized and they were working toward an October 30, 2016, full implementation date.

Crew Resource Management (CRM) and EMAS Training

The manager of flight operations training at the time of the incident was also a check airman. He had been manager of training for about 1.5 years and had been with the company for 2 years.

The airline provided three courses on CRM: new hire, captain’s upgrade, and recurrent. The new hire CRM course consisted of a 2-hour segment covering CRM background, communications processes and decision behavior, team building and leadership, workload management and situational awareness, individual factors and stress reduction, and error management. The upgrade training included 1 day of ground school in which 1 hour was dedicated to CRM. Upgrade training also incorporated a captain’s leadership course that included content on the captain’s authority, briefings, workload management, and sterile cockpit procedures in accordance with 14 CFR 121.542, “Flight Crewmember Duties.” The recurrent training included a 3.5-day ground school for captains and first officers in which 1 hour was devoted to CRM training. All courses were taught using presentation slides, open discussion, and videos created by contracted training organizations.

The captain reported after the incident that he believed he and the first officer were working well as a crew during the trip. He stated that he did not call for a transfer of controls during the landing rollout and that, in hindsight, he should have. He further mentioned that he thought it was “OK” for both crewmembers to be applying brakes. The first officer reported a “lack of communication” during the landing rollout because the captain did not say that he was taking control of the airplane. Another Eastern Air Lines first officer who had flown with the captain before the incident described the captain’s CRM as “good.”

At the time of the incident, EMAS training was not part of Eastern Air Lines’ pilot training program. The captain stated during postincident interviews that he had forgotten that an EMAS was installed at the end of runway 22, that he had read about the systems, but had not had any training on them.

FAA Oversight

The former FAA principal operations inspector (POI) stated that he had been assigned to Eastern Air Lines before the company received its operating certificate. He stated that his duties included, most critically, surveillance and reviewing the airline’s manuals, including any changes to the manuals. He traveled to the airline’s headquarters about once or twice a week.

He also stated that he interacted most with the operations management, director of safety and security, and the CEO.

The former director of safety and security stated that during his time at Eastern Air Lines, he “seldom” interacted with the FAA POI or other FAA personnel. Other management personnel stated they interacted with the FAA daily or multiple times per week, via telephone, e-mail, or in person at the FAA’s office or at Eastern Air Lines’ office. The manager of flight operations training stated that he did not directly interact with the POI and usually went through the vice president of flight operations or the chief pilot. The vice president of flight operations stated that they had been assigned a new POI 5 months before the incident and that the interaction with the new POI was “really great.”

The FAA POI at the time of the incident reported that he mostly communicated with Eastern Air Lines’ director of flight operations and chief pilot but had also communicated with the director of flight training. He categorized the communication as “very good.” He added that Eastern Air Lines was the only certificate he managed and that FAA resources were limited such that they only had one person in the office who was able to conduct checkrides in the Boeing 737. He estimated that he was at Eastern Air Lines’ operations a “couple of times a week;” however, he had not taken part in Eastern Air Lines’ pilot training. He also stated that the training in the manual for a go-around was similar to the syllabus used by other airlines, and he “assumed” that they did some go-around training in the flare and some training in low visibility. The POI stated that, following the incident, he and Eastern Air Lines management had discussed training go-arounds once the airplane was on the ground and that further discussion was needed.

Additional Information

Sterile Cockpit Regulations

The CVR also contained conversation between the flight crew during the descent and approach below 10,000 ft that was not pertinent to the flight. Title 14 CFR 121.542, “Flight Crewmember Duties” states, in part, the following:

No flight crewmember may engage in, nor may any pilot in command permit, any activity during a critical phase of flight which could distract any flight crewmember from the performance of his or her duties or which could interfere in any way with the proper conduct of those duties. Activities such as…engaging in nonessential conversations within the cockpit and nonessential communications between the cabin and cockpit crews…are not required for the safe operation of the aircraft.

…critical phases of flight include all ground operations involving taxi, takeoff and landing, and all other flight operations conducted below 10,000 feet, except cruise flight.

Runway Condition Reports from Other KLGA Arrivals

Flight crews from four flights that landed on runway 22 within 10 minutes of the incident flight reported braking as “good” or “fair.” One crew reported noticing their airplane’s antiskid brake system pulsating during the landing rollout. Others reported that there was no hydroplaning or decrease in braking performance.


Automatic terminal information service (ATIS) “Bravo” was current when the first officer, who was the pilot flying, began to brief the instrument landing system approach for runway 22. The ATIS indicated visibility 3 miles in rain, ceiling 1,500 ft broken, overcast at 2,200 ft, wind from 130º at 9 knots, and that braking action advisories were in effect. The approach briefing included the decision altitude and visibility for the approach and manual deployment of the speed brakes by the captain, with the captain stating “you’re gonna do these. I’m gonna do this” to which the first officer replied “[that] is correct.” (The airplane’s automatic speed brake module had been deactivated 2 days before the incident and deferred in accordance with the operator’s minimum equipment list, which was appropriate).

The flight crew completed the approach briefing after descending through 18,000 ft mean sea level and completed the landing checklist when the airplane was near the final approach fix.

The airplane was configured for landing with the autobrake set to 3 and the flaps set to 30º.

ATIS information “Charlie” was current at that time and indicated visibility 3 miles in rain, ceiling 900 ft broken, overcast at 1,500 ft, and wind from 120º at 9 knots.

Flight data recorder (FDR) data and postincident flight crew statements indicate that the airplane was stabilized on the approach in accordance with the operator’s procedures until the flare. The airplane crossed the runway threshold at 66 ft radio altitude at a descent rate of 750 ft per minute. When the airplane had traveled about 2,500 ft beyond the runway threshold, its descent rate decreased to near zero, and it floated during the flare. Its pitch attitude started to increase in the flare from 2.8° at a radio altitude of about 38 ft, which is high compared to the 20 ft recommended by the Boeing 737 Flight Crew Training Manual. Further, the first officer didn’t fully reduce the throttles to idle until about 16 seconds after the flare was initiated and after the airplane had touched down. The initiation of the flare at a relatively high altitude above the runway and the significant delay in the reduction of thrust resulted in the airplane floating down the runway, prompting the captain to tell the first officer to get the airplane on the ground, stating “down down down down you’re three thousand feet remaining.”

The airplane eventually touched down 4,242 ft beyond the runway threshold. According to the operator’s procedures, the touchdown zone for runway 22 was the first third of the 7,001-ftlong runway beginning at the threshold, or 2,334 ft. Touchdown zone markers and lights (the latter of which extended to 3,000 ft beyond the threshold) should have provided the flight crew a visual indication of the airplane’s distance beyond the threshold and prompted either pilot to call for a go-around but neither did. The point at which the airplane touched down left only about 2,759 ft remaining runway to stop. The airplane’s groundspeed at touchdown was 130 knots.

The captain manually deployed the speed brakes about 4.5 seconds after touchdown and after the airplane had traveled about 1,250 ft down the runway. Maximum reverse thrust was commanded about 3.5 seconds after the speed brakes were deployed, and, with fully extended speed brakes and maximum wheel brakes (which were applied at main gear touchdown) the airplane achieved increasingly effective deceleration. Its groundspeed was about 35 knots when it entered the EMAS. With the effective deceleration provided by the fully extended speed brakes, maximum wheel brakes, and reverse thrust, the flight crew would have been able to safely stop the airplane if it had touched down within the touchdown zone.

The captain later stated that he had considered calling for a go-around before touchdown but the “moment had slipped past and it was too late.” He said that “there was little time to verbalize it” and that he instructed the first officer to get the airplane on the ground rather than call for a go-around. He reported that, in hindsight, he should have called for a go-around the moment that he recognized the airplane was floating in the flare. The first officer said that he did not consider a go-around because he did not think that the situation was abnormal at that time.

Training and practice improve human performance and response time when completing complex tasks. In this case, the operator’s go-around training did not include any scenarios that addressed performing go-arounds in which pilots must decide to perform the maneuver rather than being instructed or prompted to do so. Thus, the incident flight crew lacked the training and practice making go-around decisions, which contributed to the captain’s and first officer’s failure to call for a go-around.

Following the incident, the operator incorporated go-around training scenarios in which flight crews must decide to go around rather than being instructed to do so. The company’s director of operations also stated that the company has incorporated scenarios in which go-arounds are initiated from idle power and rejected landings are performed after touchdown with the automatic speed brake inoperative. It also added a training module emphasizing that “if touchdown is predicted to be outside of the [touchdown zone], go around” and intended to require a go-around if landing outside of the touchdown zone were predicted. The operator also intended to incorporate go-around planning into the approach briefing. Flight crews would determine the cues for the touchdown zone using the airport diagram and decide at which point they would initiate a go-around if the airplane had not touched down.

Given the known wet runway conditions and airplane manufacturer and operator guidance concerning “immediate” manual deployment of the speed brakes upon landing, the captain’s manual deployment of the speed brakes was not timely. NTSB analysis of FDR data for previous landings in the incident airplane determined an average of 0.5 second for manual deployment of the speed brakes. Using the same touchdown point as in the incident, post incident simulations suggest that, if the speed brakes had been deployed 1 second after touchdown followed by maximum reverse thrust commanded within 2 seconds, the airplane would have remained on the runway surface. Therefore, the captain’s delay in manually deploying the speed brake contributed to the airplane’s runway departure into the EMAS.

During the landing roll, the captain did not announce that he was assuming airplane control, contrary to the operator’s procedures, and commanded directional control inputs that countered those commanded by the first officer. The captain later reported that he had forgotten that an EMAS was installed at the end of runway 22 and attempted to avoid the road beyond the runway’s end by applying right rudder because he thought it would be better to veer to the right. However, the first officer applied left rudder to maintain alignment with the runway centerline and to counter the airplane pulling “really hard” to the right because of the captain’s inputs. The breakdown of crew resource management during the landing roll and the captain’s failure to call for a go-around demonstrated his lack of command authority, which contributed to the incident.

At the time of the incident, EMAS training was not part of the operator’s pilot training program, but such training was added after the incident. The circumstances of this event suggest that the safety benefit of EMASs could be undermined if flight crews are not aware of their presence or purpose.


The National Transportation Safety Board determines the probable cause(s) of this incident to be:

The first officer’s failure to attain the proper touchdown point and the flight crew’s failure to call for a go-around, which resulted in the airplane landing more than halfway down the runway. Contributing to the incident were, the first officer’s initiation of the landing flare at a relatively high altitude and his delay in reducing the throttles to idle, the captain’s delay in manually deploying the speed brakes after touchdown, the captain’s lack of command authority, and a lack of robust training provided by the operator to support the flight crew’s decision-making concerning when to call for a go-around.


Aircraft Landing flare – Not specified (Factor)

Personnel issues Use of policy/procedure – Copilot (Cause)

Use of policy/procedure – Flight crew (Cause)

Lack of action – Flight crew (Cause)

Delayed action – Copilot (Factor)

Delayed action – Pilot (Factor)

Decision making/judgment – Pilot (Factor)

Organizational issues Recurrent training – Operator (Factor)

Excerpted from National Transportation Safety Board Aviation Incident Final Report  DCA17IA020


  1. Let’s go around
  2. The Organizational Influences behind the aviation accidents & incidents
  3. Speaking of going around
  4. Going around with all engines operating


minime2By Laura Duque-Arrubla, a medical doctor with postgraduate studies in Aviation Medicine, Human Factors and Aviation Safety. In the aviation field since 1988, Human Factors instructor since 1994. Follow me on facebook Living Safely with Human Error and twitter@dralaurita. Human Factors information almost every day 


The numerous safety deficiencies behind Helios Airways HCY 522 accident

Behind the well-known primary cause of this accident were less-known and less-discussed numerous regulatory and organizational factors that still can be found in many countries and airlines worldwide.


Lack of political commitment to supply the aviation authority with the resources to carry out fully its safety oversight function, safety regulatory authority not organized and staffed to effectively accomplish its regulatory and safety oversight duties, absence of leadership and oversight, inadequate work climate, operator’s incomplete management structure with qualifications of some managers not corresponding to job descriptions, airline’s philosophy and style of management not conducive to efficient and safe operations with weak work climate, large percentage of staffing with seasonal employees, reactive approach of safety management, quality assurance  not effective and deficiencies in operator’s procedures and training, are some of the factors identified.

Sounds familiar?

It’s a longish reading but worth it. So, get your favorite chair, a very big cup of Colombian coffee and enjoy!


Hellenic Republic Ministry of Transport & Communications Air Accident Investigation & Aviation Safety Board (AAIASB), Helios Airways Flight HCY522 Aircraft Accident Report. Boeing 737-31s at Grammatiko, Hellas on 14 August 2005

Published 11 / 2006

Operator: Helios Airways

Owner: Deutsche Structured Finance & Leasing Gmbh & Co

Manufacturer: Boeing Co

Aircraft Type: B 737 – 31s

Nationality: Cyprus

Registration : 5B-DBY

Place of accident: Hilly terrain in the vicinity of Grammatiko village, approximately 33 km northwest of Athens International Airport 38º 13.894’ N, 23º 58.214’ E

Date and time: 14 AUGUST 2005 – 09:03:32 h (Notes: 1. All times in the report are Coordinated Universal Time (UTC) (Local time in Hellas was UTC + 3 h). 2. Correlation of the times used in the radar and radio communication recordings, and the FDR and CVR showed differences of less than 12 seconds. The FDR time was used as the master time in this report.)

On 14 August 2005, a Boeing 737-300 aircraft, registration number 5B-DBY, operated by Helios Airways, departed Larnaca, Cyprus at 06:07 h for Prague, Czech Republic, via Athens, Hellas. The aircraft was cleared to climb to FL340 and to proceed direct to RDS VOR. As the aircraft climbed through 16 000 ft, the Captain contacted the company Operations Centre and reported a Take-off Configuration Warning and an Equipment Cooling system problem. Several communications between the Captain and the Operations Centre took place in the next eight minutes concerning the above problems and ended as the aircraft climbed through 28 900 ft. Thereafter, there was no response to radio calls to the aircraft. During the climb, at an aircraft altitude of 18 200 ft, the passenger oxygen masks deployed in the cabin. The aircraft leveled off at FL340 and continued on its programmed route.
At 07:21 h, the aircraft flew over the KEA VOR, then over the Athens International Airport, and subsequently entered the KEA VOR holding pattern at 07:38 h. At 08:24 h, during the sixth holding pattern, the Boeing 737 was intercepted by two F-16 aircraft of the Hellenic Air Force. One of the F-16 pilots observed the aircraft at close range and reported at 08:32 h that the Captain’s seat was vacant, the First Officer’s seat was occupied by someone who was slumped over the controls, the passenger oxygen masks were seen dangling and three motionless passengers were seen seated wearing oxygen masks in the cabin. No external damage or fire was noted and the aircraft was not responding to radio calls. At 08:49 h, he reported a person not wearing an oxygen mask entering the cockpit and occupying the Captain’s seat. The F-16 pilot tried to attract his attention without success. At 08:50 h, the left engine flamed out due to fuel depletion and the aircraft started descending. At 08:54 h, two MAYDAY messages were recorded on the CVR.
At 09:00 h, the right engine also flamed out at an altitude of approximately 7 100 ft. The aircraft continued descending rapidly and impacted hilly terrain at 09:03 h in the vicinity of Grammatiko village, Hellas, approximately 33 km northwest of the Athens International Airport. The 115 passengers and 6 crew members on board were fatally injured. The aircraft was destroyed.
The Air Accident Investigation and Aviation Safety Board (AAIASB) of the Hellenic
Ministry of Transport & Communications investigated the accident following ICAO practices and determined that the accident resulted from direct and latent causes.





Personnel Information

1. Captain

The Captain was male, 59 years old.

He held an Air Transport Pilot License (ATPL) issued on 27 February 1991 in accordance with JAR-FCL by LBA Germany. His ATPL license, his instrument rating category III, and his Boeing 737-300 and -800 ratings were valid until 4 June 2006.

He had attended engineering college in Dresden, East Germany from 1966 to 1970 and graduated as a pilot-engineer. He had held an ATPL issued in 1970 by the Civil Aviation Authority in East Germany.

Medical Certificate: Class A, Medical Certificate issued on 21 March 2005 and valid until 9 October 2005 with the restriction to carry two pairs of corrective lenses.

Last LPC/OPC 4 June 2005

Recurrent Training in STD 4 June 2005

Last Line Check 12 June 2005

CRM training 2 June 2005

Flying experience:

Tabla 1

The Captain had worked for the Operator for two separate time periods. According to interviews of his peers at the Operator, during the first period, he presented a typical “command” attitude and his orders to the First Officers were in command tone. During the second period, his attitude had improved as far as his communication skills were concerned.

According to an oral statement by the next of kin, the Captain was a quiet and professional pilot. His hobby was to construct and fly model aircraft. He used no drugs or medication, and he used alcohol occasionally and with moderation.

2. First Officer

The First Officer was male, 51 years old.

He held an Air Transport Pilot License (ATPL) issued in accordance with JAR-FCL by the United Kingdom. His ATPL license and Boeing 737-300 and -800 ratings were valid until 31 March 2006 and his instrument rating category III was valid until 31 October 2005.

He had attended and graduated from Chelsea College with an Engineering Diploma. He was also trained at Oxford Air Training School to become a pilot.

Medical Certificate: Class A Medical Certificate issued on 25 April 2005 and valid until 29 October 2005 with no restrictions.

Last OPC 9 March 2005

Last Line Check 3 February 2005

Recurrent training in STD 9 March 2005

CRM training 28 February 2005

Flying experience:

Tabla 2

The First Officer spent the day preceding the accident at his summer house with the family. He drove home in the evening, had a normal dinner (no alcohol) and he went to bed at about 23:00 h. He woke up early in the morning and drove to the airport in order to report for duty on time.

According to statements by his next of kin, colleagues, and friends, the First Officer was an optimist, calm, active and a social person. He had expressed his views several times about the Captain’s attitude. He had also complained about the organizational structure of the Operator, flight scheduling and he was seeking another job. He used no drugs or medication and he did not smoke or drink alcohol.

In his last three OPCs, there were the following remarks/recommendations:

9 March 2005 “Standards achieved, but with room for lots of improvement. Some difficulties met in complex tasks. Do not rush through check lists. Recommendation – improve your understanding on the use of AFS”.

3 September 2004 “Overall standard is above average. Very Good LVO recognition of abnormalities. EMERGENCY DESCENT [capital letters used by the TRE] repeated at very good standard. Keep the good work”.

13 April 2004 “Overall performance at standard – Good Manual control, and 1 ENG G/A. –Make positive control is advisory after engine failure on T/OFF not to lose direction. Repeat – OK”.

In addition, the First Officer’s training records were reviewed for the five years he worked for the Operator. The review disclosed numerous remarks and recommendations made by training and check pilots referring to checklist discipline and procedural (SOP) difficulties.

3. Cabin Attendants

There were four cabin crew members on board, all of which met Operator proficiency and medical requirements.

Cabin Attendant number four also held a UK Commercial Pilot License (JAR CPL A/IR) with an issue date of 2 October 2003, and valid until 1 October 2008. His JAA Class 1 Medical Certificate was valid from 15 July 2005 to 17 July 2006.

Medical and Pathological Information

The Captain’s samples (obtained on 18 August 2005) tested negative for major drugs of abuse, volatile poisons, and prescription and over-the-counter medications. Due to the presence of extensive burns, the determination of blood alcohol level was not possible.

The Captain’s heart muscle samples revealed the presence of minor atherosclerosis (40% obstruction) compatible with his age. A histological examination revealed the presence of recent myocardial ischaemia.

The First Officer’s samples (obtained on 15 August 2005) tested negative for carbon monoxide, volatile poisons, major drugs of abuse, and prescription and over-the-counter medications. Although ethanol was detected (34 mg/dl, or 0.034 % weight/volume – also known as blood alcohol content), the toxicological reports stated that “the time period between the death and the collection and the analysis of specimens (24 hours) may have resulted in postmortem ethanol production.” The First Officer’s heart muscle samples revealed the presence of extensive atherosclerosis (90% obstruction in the anterior descendant and circumflex coronary artery). A histological examination revealed the presence of recent myocardial ischemia.

The Department of Cardiology of the Hellenic Air Force Medical Centre predicated that “On the basis of the data that were given to us, such as the height of the flight, the fact of the existing heart function (pump function) upon crashing, and the fact that there is a similar pathologo-anatomical image both in the ‘suffering’ heart (myocardium) of the co-pilot, and in the ‘healthy heart’ of the pilot, we estimate that the brain hypoxia was the dominant and determinant cause that incapacitated the flying crew, with the findings of the heart being the matter of course and epiphenomenon of the prolonged hypoxia.”


1. Crew

Based on indubitable evidence, the Board concluded that the pressurization mode selector was in the MAN (manual) position from the time the aircraft was still on the ground and was led to believe that the selector had remained in the MAN (manual) position after the Pressure Leak Test, the last known time the particular selector had been manipulated. When the aircraft departed, the pressurization mode selector remained in the MAN (manual) position (instead of AUTO) and remained there until the aircraft impacted the ground almost three hours later. Naturally, the fact that the mode selector position was not rectified by the flight crew during the aircraft preflight preparations was crucial in the sequence of events that led to the accident.


Both the Captain and the First Officer were experienced pilots and had performed the preflight duties numerous times in the past. The Board examined the reasons why they could have made such a crucial omission.

Preflight duties included checking the Equipment Cooling switches, the Cabin Pressurization Panel, and the flight crew oxygen masks. When the pressurization mode selector is positioned to the MAN (manual) position, it is accompanied by an advisory, green light indicating MANUAL. Normally, with the mode selector on AUTO as prescribed by the Preflight Procedure, no illuminated indication should appear on the pressurization panel. Why an experienced crew would have failed to notice the presence of an indication they did not normally expect to see at this location in this phase of flight?

Why an experienced crew would have failed to notice the presence of an indication they did not normally expect to see at this location in this phase of flight?

A typical Preflight Procedure may contain between 40 and 80 actions to be performed by the First Officer, often under the pressure of the impending departure, and in the presence of a Captain who is waiting to call for the ensuing checklist. This procedure is performed from memory, aided by the fact that the actions are organized along the topographical location of panels in the cockpit. Memorization is beneficial for long lists of actions, but has the disadvantage that actions are performed automatically, without conscious effort and attention. This can and has, in the past, led to inadvertent omissions and other types of mistakes.

The Board was also sensitive to the fact that automatic execution of actions was very much affected by assumptions – in the case of performing a large number of verification steps, the assumption that all switches and indications were in the usual, normal for this phase of flight position. A superfluous green indication on the pressurization panel could be easily (inadvertently) overlooked when the perception was biased by the expectation that it should not be present.

Exacerbating this tendency (expectation bias) is the rarity with which switches (especially, and directly relevant to this case, the pressurization mode selector) are in other-than-their-normal position. A pilot automatically performing lengthy verification steps, such as those during preflight, is vulnerable to inadvertently falsely verifying the position of a switch to its expected, usual position (i.e. the pressurization mode selector to the expected AUTO position) – especially when the mode selector is rarely positioned to settings other than AUTO.

The Board was concerned that the overhead panel design was not conducive to safeguarding against these types of inadvertent omissions. Specifically, the color of the illuminated indication (green) does not typically imply something out of the ordinary, as did the amber (caution) or red (warning) – which would have likely attracted the flight crew’s attention that something was out of the ordinary.

After the Preflight Procedure, the crew was expected to orally execute a Preflight Checklist. Per the carrier’s FCOM, this checklist included a check of the pressurization panel:

[item 12 of 25] AIR COND & PRESS ………___PACK(S), BLEEDS ON, SET

The flight crew failed to detect the improper configuration of the pressurization panel during this checklist. Both the Captain and First Officer had repeatedly accomplished this checklist on many flights during their long careers. Their failure to properly accomplish the above checklist prevented them from capturing their earlier mistake. This was the first of two missed opportunities to notice and correct an earlier error. Various factors could have contributed to this failure.

The Board first examined the design of the checklist, and specifically the fact that the challenge part of this action item (“Air Cond & Press”) essentially combined two separate systems (air conditioning and pressurization). While this combination was certainly not random (the two systems used engine bleed air as an energy source), the corresponding response portion of the action item (“Pack(s), Bleeds ON, SET”) contained three different confirmations, only the third of which referred to the pressurization panel. In turn, this third confirmation referred to eight different actions – those that were performed earlier, during the Preflight Procedure. Contrary to the manufacturer’s original intention, however, many pilots informally reported that when performing the checklist and responding “SET” to the pressurization panel, they really only checked that the landing and cruise altitudes had been correctly set in the corresponding indicators.

The performance of checklists in routine, daily flight operations was also examined. In general, checklist items are performed by referencing a printed card. Like procedures, because they are performed repeatedly on the line, they are also performed by memory, typically in time-pressured circumstances (i.e. indirect pressure to maintain on-time departures). For these two reasons, checklists are often performed in a hurried, automatic fashion. From a human factors standpoint, rushing is known to lead to the inadequate allocation of attention to the task at hand – and thus to errors. Furthermore, like procedures, checklists are also vulnerable to “looking without seeing” because they are biased by the assumption that since each item verified an action performed only moments ago, then it must be already in the desired position/set.

Following takeoff, the flight crew was to perform an After Takeoff checklist, the first item of which was to check the pressurization system again and verify its settings. Although this checklist would have directed the flight crew’s attention to the pressurization panel, there was no evidence that the incorrect position of the pressurization mode selector was rectified. This was the second missed opportunity to note and correct an earlier error.

After Takeoff checklist is also usually performed under even more time pressured conditions and at a time when the pilots’ attention is consumed by other, concurrent tasks (e.g. retracting the landing gear and flaps, monitoring the climb, and communicating with ATC). The management of multiple concurrent tasks requires the division of attention resources and is known to force a person to devote insufficient attention to any one of the many tasks.

At an aircraft altitude of about 12 000 ft the cabin altitude warning horn sounded. Eight seconds later, the FDR showed the autopilot being disengaged, and re-engaged four seconds later. Eight seconds later, the FDR showed the auto-throttle being disengaged and the throttles retarded, but like the autopilot it also was re-engaged nine seconds later. Three seconds later, the No.2 radio was used to contact the Helios Airways Dispatcher.

The Board examined the flight crew’s actions to disengage the autopilot and auto-throttle, and to retard the throttles upon onset of the warning horn. Given that the expected reaction to a cabin altitude warning horn would have been to stop the climb (there was no evidence to this effect), the Board considered such actions to signify that the flight crew reacted to the warning horn as if it had been a Takeoff Configuration Warning (the two failures use the same warning horn sound). Similar occurrences had been reported by flight crews worldwide in the past.

Various factors for creating the potential confusion of the two experienced pilots were considered:

  1. In the course of his career, a pilot is generally likely to only hear the warning horn when it is associated with a takeoff and a takeoff configuration problem and most pilots are not very likely to experience a cabin pressurization problem and the associated warning horn at any time during their line flying.
  2. Stress, such as that caused by the onset of a loud, distracting alarm in the cockpit, combined with the element of surprise, is known to lead to automatic reactions. Automatic reactions, in turn, are typically those that result from experience and frequency of encounter and are therefore not always appropriate. The Board considered that the flight crew may have automatically reverted to a reaction based on memory before consciously processing the source and significance of the stress factor. This would also explain why the flight crew failed to realize the improbability of their interpretation of the horn as a takeoff configuration horn and why they failed to move on to gathering information for a new, correct diagnosis of the problem at hand. It is important to note that at no time during this sequence of events was the cabin altitude warning horn canceled.
  3. According to FDR data, at an aircraft altitude of about 17 000 ft, the MASTER CAUTION light was activated and was not canceled for 53 seconds. Two different events occurred at about this time, either one of which would have triggered the MASTER CAUTION light with the accompanying OVERHEAD indication on the Annunciator Panel to draw the attention of the pilots to a situation indicated on the Overhead Panel. The equipment cooling low flow detectors reacted to the decreased air density and one or both of the Equipment Cooling lights illuminated on the Overhead Panel. In addition, the oxygen masks deployed in the passenger cabin, illuminating the PASS OXY ON light, located further aft on the Overhead Panel. The Board was unable to determine which event occurred first and triggered the MASTER CAUTION. However, the fact that the flight crew had not canceled the first MASTER CAUTION meant that the second event did not trigger a second MASTER CAUTION as it was already on. Consequently, there was nothing to prompt the flight crew to look for a second indication on the Overhead Panel.
  4. At the time of onset of the MASTER CAUTION (and the OVERHEAD indication), workload in the cockpit was already high.
  5. Language difficulties between the Captain and the Helios Operations Centre, probably due to the fact that the Captain spoke with a German accent and could not be understood by the British engineer prolonged resolution of the problem, while the aircraft continued to climb. Moreover, the communication difficulties could also have been compounded by the onset of the initial effects of hypoxia.
  6. The Board recognized that from a human factors standpoint, preoccupation with one task (i.e. trouble-shooting the source of the Equipment Cooling problem) at the expense of another (i.e. trouble-shooting the source of the warning horn) was entirely plausible and has happened to experienced pilots.
  7. The combination of hypoxia and distractions generally increases stress levels. Stress is known to render human cognition (e.g. memory, attention, decision-making, risk management, communication skills) particularly vulnerable to errors
  8. The flight crew of HCY522 did not exhibit adequate CRM to help overcome the individual errors and to detect a dangerous situation that deteriorated as the aircraft continued to climb.

Given the ongoing distractions, the Captain, at least, may never have consciously and fully registered the onset of the indications and/or their significance. Unfortunately, although there were partial data to somewhat deduce the Captain’s actions at this time (from his communication exchanges with the Operator’s dispatcher and engineer), there was no possibility to establish the First Officer’s actions during this same time.

The Board evaluated what the cabin crew’s reactions might have been when the aircraft continued to climb and there was no announcement from the flight deck. There was no Operator procedure to address such a contingency. As emphasized by the Cabin Crew Manager in his post-accident statement, however, cabin crews were encouraged to take initiative. The Manager expressed his conviction that the particular cabin crew was well trained and by nature fully bound to have taken the initiative to seek an explanation for the unusual situation they were facing. The Board considered the fact that even if this was the case, it was hard for a cabin crew to assess how long to wait before contacting the flight deck – and in this case, time was of the essence as the hypoxia effects grew increasingly stronger. It was not possible to determine whether any of the cabin crew members attempted to contact the flight crew or enter the flight deck.

Data from the CVR only contained to the last 30 minutes of the accident flight and showed that at least one cabin crew member retained his consciousness for the duration of the flight and entered the flight deck more than two hours after takeoff. At the beginning of the climb phase, this cabin attendant was likely seated next to the aft galley.

In order for him to have moved forward in the aircraft to reach the flight deck, he must have used a portable oxygen bottle.

The Board found the fact that this cabin attendant might not have attempted to enter the flight deck until hours after the first indication that the aircraft was experiencing a nonnormal situation quite puzzling. Of course, in the absence of a longer-duration CVR, it was not possible to know whether this or any other cabin crew member had attempted to or succeeded in entering the flight deck. From the sounds recorded on the CVR, however, the Board could ascertain that this cabin attendant entered the cockpit using the emergency access code to open a locked cockpit door.

2. Operator

2.1 Maintenance

Based on evidence from the Helios Airways Technical Department documents relevant to manpower planning, the front line maintenance task force group consisting of four to five licensed engineers and two to three mechanics, changed, as far as the individual persons of the first group were concerned, by more than 80 % three times within 16 months (oldest EMPLOYMENT DATE: 01/11/2003 – first END DATE: 04/03/2005). The longest stay with Helios Airways up to 14 August 2005 was 21 months and the shortest three days. Both example cases above were licensed engineers and categorized as “Permanent” in the column “Employment Status” of the document. The same column contained another category, “Contract”, reserved for those licensed engineers hired through employment agencies.

Between November 2003 and August 2005 (one week before the accident occurred), 13 licensed engineers (six different European nationalities) were employed by the Helios maintenance department and subsequently left Helios. Six of them were contracted, which meant that they were paid by the employment agency that placed them with the airline. The Board believed that the very high turnover rate of maintenance personnel was not conducive to establishing and maintaining a sense of continuity and teamwork among employs, and this probably worked against setting a good foundation for proactive management and resolution of any issues in the maintenance department.

This situation has been raised as a Non-Conformance Report (NCR) during an audit carried o April 2005. The certifying staff level (number of licensed engineers), not including the Maintenance Manager, was annotated in the NCR as insufficient to meet the requirements of Part 145. The manpower plan appeared to only be a guide and did not fully reflect the current status of manpower usage or requirements. Despite the corrective action to the NCR stated by the Operator’s maintenance management to improve the situation, the responses by airline management continued to prove inadequate to provide the necessary resources and financial support.

The former Technical Manager of the Operator was asked why he resigned in January 2005 after having served the company for more than four years. He answered that the reason for his decision was the mismanagement in cases such as:

a) Staffing of key posts e.g. Quality Manager, Flight Operations Manager, with individuals who either did not have the required qualifications by the Operator’s Policy prerequisites, or did not possess managerial competence;

b) Lack of business planning;

c) Incoherent corporate operations; and

d) Occasional coverage of personnel requirements in all specialties of the corporate operations.

2.2. Crew scheduling

According to the records made available to the Board, the crew duty times were within limits and followed the prescribed standards. However, in view of these records and a number of statements made to the Board, it had reservations on this subject, given that the records submitted required extensive examination to validate flight and duty times for the flight crew. The Board noted that inspectors/auditors in previous audits had annotated comments that the Captain’s Deviation Reports (CDRs) showed flight and duty times that exceeded the approved limits and were not recorded or reported to the DCA.

The Board also noted statements that the scheduling of flights was based on unrealistic flight times for some routes in order to ensure flight planned adherence to flight time limitations which subsequently were exceeded

2.3. Crew Training

According to the Helios Flight Training Manual, the simulator training syllabus included rapid decompression situations, but not gradual decompression (slow loss of pressurization) situations. Consequently, the flight crews were likely not sensitized to monitoring and detecting a more insidious, gradual loss of pressurization situation.

The Board identified a specific requirement for training of both flight and cabin crews on the phenomena associated with hypoxia. However, based on witness statements, the Board was led to believe that this requirement was not fulfilled in practice but remained a requirement “on paper.” The Board noted that this situation was not unique to Helios Airways, because the lack of hypoxia training to sensitize flight crews to detecting an insidious gradual decompression or non-pressurization of the aircraft during climb, was a common situation in the airline industry.

Interviews with a number of cabin crew members (including Cabin Chiefs) revealed a number of deficiencies. In particular, cabin crews appeared confused and responded differently to questions that concerned the number and type of oxygen masks on the B737 flight deck, the availability and exact procedure of means available to open the cockpit door, and whether passenger oxygen masks provided breathable oxygen at high altitude.

Furthermore, deficiencies were also identified in the Operator’s procedures that prescribed actions to be taken in the event that, after passenger oxygen mask activation, the aircraft did not begin to descend or at least to level-off. However, it was also determined that other airlines in Cyprus and in Greece did not have such procedures documented in their manuals.

The Board found training deficiencies and inconsistencies. Although it was determined that some of these issues were probably not implicated in the accident, some aspects of the procedures determined at Helios could be considered unsafe.

3. Organizational Issues

The management structure at Helios Airways at the time of the accident was incomplete, notably the position of the Manager Training Standards. The Flight Operations Manager had assumed the responsibilities of the Manager Training Standards, pending a reorganization of the Operations Division and the arrival of the new Chief Operating Officer at the beginning of August 2005. Furthermore, the Chief Pilot was in a position to deputize for the Training Manager Standards. The qualifications of some of the interviewed managers did not correspond to the qualifications listed in their job descriptions. These deficiencies in management may have been related to the failure of the Operator to recognize and take appropriate corrective actions to remedy the chronic checklist and SOP omissions exhibited by the First Officer and documented in his training records.

The Accountable Manager was characterized as unapproachable, with little regard or concern for safety or for the well-being of the company employees, and whose only interest was the profitability of the Operator.

The Board acquired the sense that the overall philosophy and style of management at Helios Airways was not conducive to efficient and safe operations. This impression was corroborated by the UK inspector’s comments in July of 2004 expressing concern about the potential that flight safety was being compromised due to “the lack of operational management control” and the hesitancy with which some improvements were made, were noted by another inspector a year later.

The Board considered potential implications of the multi-national staff composition at Helios Airways and how they might have affected the safety of flight operations. Multinational teams often led to a weak work climate because people of different cultural groups operated based on a set of values and perceptions unique to their common historical/social/geographical background. These types of differences might lead to communication and collaboration problems.

Another area of concern that arose from the composition of Helios Airways staff stemmed from the large percentage (33%) of staffing with seasonal (part-time) employees. Naturally, this was expected for companies whose operations mainly catered to the tourist industry and were, by definition, seasonal. The short-term hiring of pilots and engineers when the operational tempo and demands were significantly higher in the spring and summer allowed the airline to maintain a skeleton staff to cover the less loaded winter months. Insofar it affected work climate, however, frequent changes in staff composition could be detrimental to the development of professional and personal ties, and did not promote the required level of comfort among employees, and among employees and management, particularly with respect to the submission and discussion of incidents and problems. Employees lacked a sense of continuity, both for their own job as well as that of their colleagues, and cockpit and cabin crew did not have the opportunity to develop operational experience together in various routine and non-routine situations. Employees, finally, did not develop a feeling of ownership and responsibility towards operations and the Operator.

Provisions existed in manuals for an accident prevention/safety management program at Helios Airways. However, it was not at all clear whether the Operator adhered to the standards set forth in the relevant publications. Furthermore, these standards seemed to promote a reactive approach rather than emphasizing the benefits of a more effective, proactive stance to safety management. More important, the standards did not clearly and definitively outline the role and responsibility of management (a key element in any safety management program) in ensuring and maintaining safe operations of the company. The Board found reasons for further concern in the statement by the Chief Operating Officer of Helios Airways. By referring to tight schedules both for employees and aircraft utilization, the Chief Operating Officer appeared to suggest that both resources were utilized to the limits. The Board noted that tight scheduling, work under time pressure and considerable amounts of overtime work were not conducive to maintaining a safe work environment. These conditions were likely a fertile basis for human factor errors in flight operations and aircraft maintenance.

Management pilots appeared to be insufficiently involved in their managerial duties, this led the Board to note that the Operator lacked the mechanism and means to sufficiently and correctly monitor its pilots and to take decisive and corrective action when and as necessary.

Training and duty records were found to be incomplete, with no evidence of any type of a follow-up.

Manuals were found to be in part deficient; they did not always adhere to regulations, and on some issues they were out of date. This suggested that an underlying pressure was prevalent to proceed with little regard for the required formalities (which often equaled an assurance for safety).

Lastly, the Board also reviewed the actions of the Ground Engineer team that conducted maintenance on the aircraft prior to its departure, so as to form an opinion about the operation of the maintenance department at Helios Airways as a whole. The inexplicable inconsistencies in the actions that were or were not performed, the actions recorded, and the actions described as having been performed by Ground Engineer No. 1 on the morning of 14 August 2005 were considered by the Board to confirm the idea that the Operator was not effectively promoting and maintaining basic elements of safety in its culture.

4. Department of Civil Aviation in the Republic of Cyprus

At the time of the accident, the Safety Regulation Unit (SRU) was diachronically not organized and staffed to effectively accomplish its regulatory and safety oversight duties. The main problems that characterized the Unit and each of its three Sections (Operations, Airworthiness, and Licensing) already back in 1999 (the time of the first available audit report) appeared to still persist to this day, as evidenced by the various evaluation reports reviewed. The number of employed personnel was insufficient in relation to the actual workload. The mission and strategy of each Section, including its processes and standard operating procedures, appeared not to be officially laid out in writing. Selection and training criteria and resources, as well as detailed job descriptions, were not available. By extension, the qualifications, training, and hands-on expertise of most employees were probably inadequate. Vital positions (e.g. Head of the Operations Section) remained vacant. Some key functions (e.g. issuance and validation of air transport pilot licenses; issuance and record-keeping of medical certificates) were not performed. Other key functions (e.g. inspections) were possibly not accomplished per schedule because qualified personnel was not readily available and external resources had to be relied on.

This diachronic absence of leadership and oversight both across and within the three Sections presented a major obstacle that hindered the effective work of any one of the Sections. The resulting work climate within the SRU was not conducive to good performance even by qualified personnel; this became apparent in the nature of the oral statements given by the employees that included charges and complaints, as well as direct accusations and finger-pointing. Given the situation within the SRU, it was probably difficult for the Unit to instill a level of esteem from the aviation industry, and specifically in the areas and activities that it was tasked to regulate and oversee.

To accomplish its safety oversight duties, the Unit relied heavily on the UK CAA to furnish (based on a contractual agreement) inspectors to carry out the ICAO and EU required inspections. Based on the contractual agreements between the Cyprus DCA and the UK CAA, the role of the latter was undoubtedly intended to be advisory in nature. In reality, however, the DCA appeared to have been fostering and maintaining a relationship of complete dependence on the UK CAA, and, in most cases, appeared to be simply accepting its services without questioning them and without making an effort to assume ownership and thus build on them. The Board was particularly concerned to find that almost all of the Operator’s audit reports until about the time of the accident were signed by the UK CAA inspectors without any comments and/or a signature by an employee of the Cyprus DCA. Where the situation at the audited Operator seemed to repeatedly yield deficiencies and issues that required often urgent attention, the Board found no evidence that the DCA would actually “step in” and take action to ensure that the Operator complied and took corrective actions and, consequently, was safe and legal to continue its flight operations. As mentioned in the evaluation by a private firm in 2005 “The UK CAA representatives acknowledged that their current role in Cyprus is as advisors. However, this remains unclear since the existing contracts indicate, and records confirmed, that the UK CAA inspectors exercised a more direct, “hands-on” approach.”

The relationship of dependence was also evident from the evaluation by the private firm which found that the DCA had not taken ownership of documents prepared by the UK CAA and which described the internal operations of the Flight Operations and the Airworthiness Sections of the SRU. The SRU appeared to have adopted the manuals without completing missing sections and/or tailoring them to their needs, or trained inspection personnel to use them.

In trying to explain the reasons behind the slow progress in strengthening the DCA capabilities, the Board considered the role of Governmental support and how that may have been affecting the DCA’s ability to evolve and better embrace its safety oversight responsibilities. The Board noted that the 2002 ICAO audit clearly attributed at least part of the situation to the fact that the DCA operated as a functional department of the Ministry of Communications and Works. The 2005 European Commission evaluation directly faulted the absence of the necessary “… political commitment [of the Cypriot Government] to supply this Department [DCA] with the resources to carry out fully its safety oversight function and to reorganize the chain of command in order to give safety the high priority it deserves inside the organization.”

What became apparent from the Board’s consideration of the situation at the Cyprus DCA, and what was evident from the review of the audits/evaluations of DCA, was that the DCA, and the SRU in particular, lacked the required expertise to move forward, become independent, and fulfill the international obligations of Cyprus as contained in the Chicago Convention and its Annexes. Despite numerous action plans since 1999 to ensure the availability of properly trained and qualified inspectors, there were no tangible indications of progress.

A review of the audits and follow up audits of Cyprus DCA performed by ICAO, EASA and JAA, disclosed several important findings, which should have been actioned in the shortest possible time. No records were obtained that would have documented any remedial action considered, initiated or completed. It was of concern to the Board that there was no evidence of actions and enforcement by the international regulatory agencies to require timely implementation of an acceptable action plan, although they had clearly established that Cyprus’ international obligations were not being met.


1. Findings

1.1 Flight Crew

  1. The flight crew was licensed and qualified for the flight in accordance with applicable regulations.
  2. The flight crew held valid medical certificates and was medically fit to operate the flight.
  3. Although atherosclerosis was found (minor atherosclerosis for the Captain and extensive atherosclerosis for the First Officer), the Hellenic Air Force Aviation Medical Centre estimated that brain hypoxia was the dominant and determinant cause of incapacitation.
  4. The flight crew was adequately rested and their flight and duty times were in compliance with Cyprus DCA and Operator requirements.
  5. During the Preflight procedure, the Before Start and the After Takeoff checklists completion, the flight crew did not recognize and correct the incorrect position of the pressurization mode selector (MAN position instead of AUTO).
  6. The green light indication that the pressurization mode selector was in MAN (manual) position should have been perceived by the flight crew during preflight, takeoff, and climb.
  7. At an aircraft altitude of 12 040 ft and at a cabin pressure that corresponds to an altitude of 10 000 ft, about 5 minutes after takeoff, the Cabin Altitude Warning horn sounded.
  8. The initial actions by the flight crew to disconnect the autopilot, to retard and then again advance the throttles, indicated that it interpreted the warning horn as a Takeoff Configuration Warning.
  9. The incorrect interpretation of the reason for the warning horn indicated that the flight crew was not aware of the inadequate pressurization of the aircraft.
  10. There were numerous remarks in the last five years by training and check pilots on file for the First Officer referring to checklist discipline and procedural (SOP) difficulties.
  11. The flight crew contacted the company Operations Centre Dispatcher and referred to a Takeoff Configuration Warning horn and the Equipment Cooling lights.
  12. Communications between the flight crew and the company Operations Centre Dispatcher were not recorded; nor was there a regulatory requirement to record such communications.
  13. At an aircraft altitude of 17 000 to 18 000 ft, the Master Caution was activated and was not canceled for 53 seconds. The reason for its activation may have been either the inadequate cooling of the Equipment or the deployment of the oxygen masks in the cabin. The activation for either of the above two reasons does not permit identification of the other reason. Independently of the Master Caution indication, there are separate indications for both malfunctions on the overhead panel.
  14. The flight crew possibly identified the reason for the Master Caution to be only the inadequate cooling of the Equipment that was indicated on the overhead panel and did not identify the second reason for its activation, i.e., passenger oxygen masks deployment, that was later also indicated on the overhead panel. The crew became preoccupied with the Equipment Cooling fan situation and did not detect the problem with the pressurization system.
  15. The workload in the cockpit during the climb was already high and was exacerbated by the loud warning horn that the flight crew did not cancel.
  16. The remarks and observations by training pilots and check pilots with respect to the First Officer’s performance explained the omissions of the flight crew in its performance of the Preflight procedures, the Before Start and the After Takeoff checklists, as well as the non-identification of the warnings and reasons for the activations of the warnings on the flight deck during the climb to cruise.
  17. Before hypoxia began to affect the flight crew’s performance, inadequate CRM contributed to the failure to diagnose the pressurization problem.
  18. The flight crew probably lost useful consciousness as a result of hypoxia sometime after their last radio communication on the company frequency at 06:20:21 h, approximately 13 minutes after takeoff.
  19. Histological examinations revealed the presence of recent myocardial ischemia in both pilots, which according to the Hellenic Air Force Aviation Medical Centre (KAI) was likely due to the extended exposure to hypoxia.
  20. The toxicology test measured ethanol (34 mg/dl or 0.034 % weight/volume) in the specimen of the First Officer. The toxicological report stated that in view of the conditions, the finding may have resulted from postmortem ethanol production.

1.2. Cabin Crew

  1. The cabin crew members were trained and qualified in accordance with existing regulations.
  2. The cabin crew members were adequately rested and their duty times were in accordance with existing regulations.
  3. After the deployment of the oxygen masks in the cabin, the cabin crew members would have expected initiation of a descent or at least leveling-off of the aircraft.
  4. It could not be determined what actions were taken by the cabin crew members after deployment of the oxygen masks in the cabin, nor whether any of the cabin crew members attempted to contact the flight crew or enter the flight deck after passenger oxygen masks deployment.
  5. Shortly before flame out of the left engine, a member of the cabin crew was observed by an F-16 pilot to enter the flight deck, to sit in the captain’s seat, and to attempt to gain control of the aircraft.
  6. The above cabin crew member held a Commercial Pilot License.

1.3 Aircraft

  1. The aircraft held a valid Certificate of Airworthiness.
  2. The mass and centre of gravity of the aircraft were within prescribed limits.
  3. The aircraft had been supplied with the required amount of fuel. Fuel was not a factor in this accident.
  4. No deferred maintenance defects had been recorded.
  5. Data retrieved from the non-volatile memory (NVM) of the No. 2 cabin pressurization controller for at least the last 42 flights revealed a pressurization leak or insufficient inflow of air for reasons that could not be determined.
  6. There were nine write-ups related to the Equipment Cooling system in the Aircraft Technical Log from 9 June to 13 August 2005.
  7. The maintenance actions performed in the early morning hours of the day of the accident comprised:
    • A visual inspection of the rear right door (R2), no defects were found;
    • A pressurization test, no leakage was found.
  1. The record of the maintenance actions in the Aircraft Technical Log was incomplete.
  2. After the pressurization test, the pressurization mode selector was not selected to AUTO. Although not a formal omission, it would have been prudent to position the pressurization mode selector back to AUTO.
  3. The first recorded data of the accident flight on the non-volatile memory (NVM) chip in the cabin pressurization controller was at 10 000 ft cabin altitude (12 040 ft aircraft altitude). The data showed that the pressurization system was operating in the manual mode.
  4. The aircraft departed the holding pattern and started descending from FL340 when the left engine flamed out from fuel depletion. The right engine also flamed out from fuel depletion shortly before impact.
  5. The aircraft was structurally intact before impact.
  6. The aircraft was destroyed by the impact.

1.4 Manufacturer

  1. The description in the Boeing AMM for the procedure for the pressurization check (under the heading “Put the Airplane Back to its Initial Condition”) was vague. It did not specify an action item that the pressurization mode selector be returned to the AUTO position after the pressurization check.
  2. The manufacturer’s Preflight procedure and checklists (Before Start and After Takeoff) for checking and verifying the position of controls on the pressurization panel were not consistent with good Human Factors principles and were insufficient to guard against omissions by flight crews.
  3. The manufacturer’s procedures should have contained enough redundancy to ensure that the pressurization system was properly configured for flight. Because the position of the pressurization mode selector was critical for pressurization, the specific action should have been explicitly listed in the checklists referring to the pressurization system (Before Start and After Takeoff).
  1. The use of the same aural warning to signify two different situations (Takeoff Configuration and Cabin Altitude) was not consistent with good Human Factors principles.
  2. Over the past several years, numerous incidents had been reported involving confusion between the Takeoff Configuration Warning and Cabin Altitude Warning on the Boeing 737 and NASA’s ASRS office had alerted the manufacturer and the aviation industry.
  3. Numerous incidents had been reported worldwide involving cabin pressurization problems on the Boeing 737. A number of remedial actions had been taken by the manufacturer since 2000, but the measures taken had been inadequate and ineffective in preventing further similar incidents and accidents.

1.5. ATC

  1. The air traffic controllers in Nicosia and Athens, who handled flight HCY 522 were properly licensed and properly qualified.
  2. The ATC facilities in Nicosia and Athens were appropriately staffed and the communication equipment operated per regulations. There were no communications or navigational aid abnormalities.
  3. Nicosia ACC informed by telephone Athinai ACC that flight HCY 522 was not responding to its radio calls while approaching EVENO, but did not use the formal ICAO procedure (Doc 4444) for the two-way Radio Communication Failure (RCF).
  4. One minute before the flight entered the Athinai FIR, the Athinai ACC controller “accepted” the flight, but did not seek communication with it when it entered the FIR and failed to contact Athinai ACC as prescribed.
  5. The above-mentioned actions by Nicosia and Athinai ACCs did not contribute to the formation of events of the accident.

1.6. EASA, JAA, and ICAO

  1. Despite several EASA, JAA and ICAO audit and follow up audit findings performed on Cyprus DCA, there was no enforcement of implementation of action plans in order to meet its international obligations in the shortest possible time.

1.7. Flight HCY522

  1. When the flight HCY522 was intercepted by the F-16s, the F-16 lead pilot reported that there was no visible damage to the Boeing 737 aircraft, that the Captain’s seat was vacant, the person in the First Officer’s seat was not wearing an oxygen mask and was slumped over the controls, and some seated passengers in the cabin were observed wearing oxygen masks.
  2. Shortly before the aircraft started descending, the F-16 pilot reported that a man wearing clothing of a specific color entered the cockpit and sat down in the Captain’s vacant seat. He did not appear to be wearing an oxygen mask. He seemed to make efforts to gain control of the aircraft. It was determined that this man was a cabin attendant who held a Commercial Pilot License.
  1. When the left engine flamed out due to fuel depletion, the aircraft exited the holding pattern and started a left descending turn, and followed an uneven flight path of fluctuating speeds and altitudes. Shortly before impact, the right engine also flamed out from fuel depletion.
  2. The cabin crew member in the cockpit attempted to transmit a MAYDAY message, which was recorded on the CVR. However, the MAYDAY calls were not transmitted over the VHF radio because the microphone key, as shown by the FDR, was not pressed. The performance of the cabin crew member was very likely impaired by the hypoxic and stressful conditions.
  3. Three of the four portable oxygen cylinders on board the aircraft had most likely been used.
  4. The cabin altitude was calculated to have been about 24 000 ft, while the aircraft was at a cruise level of 34 000 ft (FL340).
  5. The duration (30 minutes) of the CVR installed on the aircraft was insufficient to provide key information that would have clarified the chain of events during the climb phase of the flight. The CVR stopped recording when the engines flamed out.

1.8. Operator

  1. The After Takeoff checklist section referring to the pressurization system in the Operator’s QRH had not been updated according to the latest Boeing revision.
  2. The manuals, procedures, and training of the Operator, and to a large extent of the international aviation industry, did not address the actions required of cabin crew members when the passenger oxygen masks have deployed in the cabin and, during climb to cruise, the aircraft has not start descending or at least leveled off, and no relevant announcement has been made from the flight deck.
  3. The absence of applied hypoxia training at the Operator, and to a large extent at other airlines, for airline transport pilots, increased the risk of accidents because of the insidious nature of incapacitation during climb to cruising altitude as a result of pressurization anomalies or gradual loss of pressurization.
  4. There were organizational safety deficiencies within the Operator’s management structure and safety culture as evidenced by diachronic findings in the audits prior to the accident, including:

a) Inadequate Quality System;

b) Inadequate Operational Management control;

c) Inadequate Quality and Operations Manual;

d) Cases of non-attendance of management personnel at quarterly management quality review meeting, as required;

e) Organization, management, and associated operational supervision not properly matched to the scale and scope of operations;

f) Inadequate monitoring of pilot certificates and training;

g) Insufficient involvement of management pilots in managerial duties, due to lack of time;

h) Incompletely updated training and duty records;

i) Lack of updating of some manuals and in part not fully in compliance with regulations;

j) Key management personnel at time performing the work of two positions;

k) Periods of vacant key management positions;

l) Inadequate remedial actions on audit findings, including level one findings, which could cause suspension of the AOC.

1.9. Cyprus DCA

  1. Organizational safety related deficiencies existed within the Cyprus DCA from at least 1999 and continued to the time of the accident, although some corrective actions were exercised since 2003. These deficiencies prevented the DCA from carrying out its safety oversight obligations within Cyprus, as evidenced by findings in previous audits, including:

a) Lack of resources and qualified personnel, and inability to adequately perform the safety oversight activities as required by ICAO;

b) Over-reliance on the UK CAA;

c) Inadequate on-the-job training for Cypriot inspectors to assume the duties for the DCA;

d) Lack of DCA internal expertise to assess the effectiveness or the technical aspects of the UK CAA inspections and the work performed;

e) Ineffectiveness of the DCA in bringing the Cyprus Civil Aviation legislation and regulations into compliance with the international requirements (ICAO Standards and Recommended Practices);

f) Inadequacy of the structure of the DCA to support safety oversight on current and future operations under the present circumstances;

g) No risk management process;

h) Non-exploitation by the DCA of the full scope of contracted services from the UK CAA, related to on-the-job training of Cyprus Flight Inspectors for reasons beyond the control of the UK CAA;

i) Non-assumption of responsibility of the DCA in directing the UK CAA regarding the accomplishment of its contractual duties;

j) Lack of effective implementation of the corrective action plans from previous audits (ICAO – 46.57 % non-implementation, when an excess of 15% non-implementation generally indicated significant problems in terms of State oversight capability).

2. Causes

2.1 Direct Causes

  1. Non-recognition that the cabin pressurization mode selector was in the MAN (manual) position during the performance of the: a) Preflight procedure; b) Before Start checklist; and c) After Takeoff checklist.
  2. Non-identification of the warnings and the reasons for the activation of the warnings (cabin altitude warning horn, passenger oxygen masks deployment indication, Master Caution), and continuation of the climb.
  3. Incapacitation of the flight crew due to hypoxia, resulting in the continuation of the flight via the flight management computer and the autopilot, depletion of the fuel and engine flameout, and impact of the aircraft with the ground.

2.2. Latent causes

  1. The Operator’s deficiencies in organization, quality management, and safety culture, documented diachronically as findings in numerous audits.
  2. The Regulatory Authority’s diachronic inadequate execution of its oversight responsibilities to ensure the safety of operations of the airlines under its supervision and its inadequate responses to findings of deficiencies documented in numerous audits.
  3. Inadequate application of Crew Resource Management (CRM) principles by the flight crew.
  4. Ineffectiveness and inadequacy of measures taken by the manufacturer in response to previous pressurization incidents in the particular type of aircraft, both with regard to modifications to aircraft systems as well as to guidance to the crews.

2.3. Contributing Factors to the Accident

  1. The omission of returning the pressurization mode selector to AUTO after unscheduled maintenance on the aircraft.
  2. Lack of specific procedures (on an international basis) for cabin crew procedures to address the situation of loss of pressurization, passenger oxygen masks deployment, and continuation of the aircraft ascent (climb).
  3. Ineffectiveness of international aviation authorities to enforce implementation of corrective action plans after relevant audits.

Excerpted from the Hellenic Republic Ministry of Transport & Communications Air Accident Investigation & Aviation Safety Board (AAIASB), Aircraft Accident Report Helios Airways Flight HCY522.


  1. The Organizational Influences behind the aviation accidents & incidents
  2. Normalization of Deviance: when non-compliance becomes the “new normal” 
  3. Unstable approach and hard landing. Final report


minime2By Laura Victoria Duque Arrubla, a medical doctor with postgraduate studies in Aviation Medicine, Human Factors and Aviation Safety. In the aviation field since 1988, Human Factors instructor since 1994. Follow me on facebook Living Safely with Human Error and twitter@dralaurita. Human Factors information almost every day

Emirates B773 accident on 03 August 2016, Interim Statement

There were no aircraft systems or engine abnormalities up to the time of the Accident


Air Accident Investigation Sector
General Civil Aviation Authority
The United Arab Emirates
Accident – First Interim Statement –
AAIS Case No: AIFN/0008/2016
Runway Impact During Attempted Go-Around

Operator: Emirates
Make and Model: Boeing 777-31H
Nationality and Registration: The United Arab Emirates, A6-EMW
Place of Occurrence: Dubai International Airport
State of Occurrence: The United Arab Emirates
Date of Occurrence: 3 August 2016

Occurrence Brief  (See: Going around with no thrust. Emirates B773 accident at Dubai on August 3rd, 2016, interim report)
Occurrence Reference : AIFN/0008/2016
Occurrence Category : Accident
Name of the Operator : Emirates
Manufacturer : The Boeing Company
Aircraft Model : 777-31H
Engines : Two Rolls-Royce Trent 892
Nationality : The United Arab Emirates
Registration : A6-EMW
Manufacture Serial Number : 32700
Date of Manufacture : 27 March 2003
Flight Hours/Cycles : 58169/13620
Type of Flight : Scheduled Passenger
State of Occurrence : The United Arab Emirates
Place of Occurrence : Runway 12L, Dubai International Airport
Date and Time : 3 August 2016, 0837 UTC
Total Crewmembers : 18 (two flight and 16 cabin)
Total Passengers : 282
Injuries to Passengers and Crew : 30, four serious, 26 minor (The number of injured has been updated since the Preliminary Report was published)
Other Injuries : One firefighter (fatal)
Nature of Damage : The Aircraft was destroyed

Investigation Objective
This Investigation is performed pursuant to the United Arab Emirates (UAE) Federal
Act No. 20 of 1991, promulgating the Civil Aviation Law, Chapter VII ̶ Aircraft Accidents,
Article 48. It is in compliance with Part VI, Chapter 3 of Part VI, Chapter 3, of the Civil Aviation Regulations (CARs) of the United Arab Emirates, and in conformity with Annex 13 to the Convention on International Civil Aviation.
The sole objective of this Investigation is to prevent aircraft accidents and incidents.
It is not the purpose of this activity to apportion blame or liability.
This first anniversary Interim Statement gives a brief of the Investigation progress
and should be read in conjunction with the Preliminary Report number AIFN/0008/2016 that was published on 5 September 2016.

This Interim Statement is released in accordance Standard 6.6 of ICAO Annex 13
and paragraph 7.4 of UAE CAR Part VI, Chapter 3.
Later Interim Statements/Reports, or the Final Report, may contain altered
information in case of new evidence becoming available during the ongoing investigation.

Investigation Process
The occurrence was classified as an Accident and the Air Accident Investigation
Sector (AAIS) of the United Arab Emirates assigned an Accident Investigation File Number AIFN/0008/2016 for the case.
The AAIS formed the Investigation team led by the investigator-in-charge (IIC) and
members from the AAIS for the relevant investigation aspects. The National Transportation Safety Board (NTSB) of the United States, being the State of the Manufacture and Design, and the Air Accidents Investigation Branch (AAIB) of the United Kingdom, being the State of Manufacture of the engines, were notified of the Accident and both States assigned Accredited Representatives assisted by Advisers from Boeing and Rolls-Royce. In addition, the Operator assigned an Adviser to the IIC. The AAIS is leading the Investigation and will issue a Final Report.
This Interim Statement is publicly available at:

Interim Statement
This first anniversary Interim statement gives a brief account of the progress of the
Investigation into the subject Accident. The statement is released in accordance Standard 6.6 of ICAO Annex 13 and paragraph 7.4 of UAE CAR Part VI, Chapter 3.
The Accident occurred on 3 August 2016 and involved an Emirates Boeing 777-300
Aircraft, registration A6-EMW, operating a scheduled passenger flight EK521, that had
departed Trivandrum International Airport (VOTV), India, at 0506 UTC for Dubai International Airport (OMDB), the United Arab Emirates. At approximately 0837:38 UTC, the Aircraft impacted the runway during an attempted go-around at Dubai International Airport.
The Aircraft sustained substantial structural damage as a result of the impact and its
movement along the runway and was eventually destroyed by fire. Twenty-one passengers, one flight crewmember, and four cabin crewmembers sustained minor injuries. Four cabin crewmembers sustained serious injuries. Approximately nine minutes after the Aircraft came to rest, a firefighter was fatally injured as a result of the explosion of the center wing fuel tank.
Regarding the operation of the flight the Investigation is working to determine and
analyze the human performance factors that influenced flight crew actions during the landing and attempted go-around.
In addition, the Investigation has reviewed and has identified safety enhancements
related to the validity of weather information that was passed to the flight crew, and
communication between air traffic control and the flight crew.
A detailed examination was performed of the Aircraft evacuation systems, including
the operation of emergency escape slides in a non-normal aircraft resting position, and the effects of wind on the escape slides.
A large number of aircraft systems were tested with the assistance of the manufacturers and analysis of the data downloaded indicates that there were no Aircraft systems or engine abnormalities up to the time of the Accident.

The UAE GCAA Air Accident Investigation Sector continues to collaborate with the
State authorities and other organizations involved in areas of interest including flight
operations, human performance, training standards, procedures, aircraft systems, passenger evacuation and airport emergency response.

This Interim Statement is issued by:
The Air Accident Investigation Sector
General Civil Aviation Authority
The United Arab Emirates
P.O. BOX 6558, Abu Dhabi.
Email: ACCID@gcaa.gov.ae


1. United Arab Emirates, General Civil Aviation Authority, Air Accident Investigation Sector. Accident – First Interim Statement -AAIS Case No: AIFN/0008/2016


  1.  Going around with no thrust. Emirates B773 accident at Dubai on August 3rd, 2016, interim report
  2. When the error comes from an expert: The Limits of Expertise
  3. Let’s go around
  4. Speaking of going around
  5. Going around with all engines operating
  6. Multitasking in Complex Operations, a real danger
  7. The Organizational Influences behind the aviation accidents & incidents 


minime2By Laura Duque-Arrubla, a medical doctor with postgraduate studies in Aviation Medicine, Human Factors and Aviation Safety. In the aviation field since 1988, Human Factors instructor since 1994. Follow me on facebook Living Safely with Human Error and twitter@dralaurita. Human Factors information almost every day 

Out of line

Some personal issues will continue to keep me out of line for a while. Hope they resolve completely in no long time so I’d be able to publish again soon. Thanks a lot for reading me.


minime2By Laura Victoria Duque Arrubla, a medical doctor with postgraduate studies in Aviation Medicine, Human Factors and Aviation Safety. In the aviation field since 1988, Human Factors instructor since 1994. Follow me on facebook Living Safely with Human Error and twitter@dralaurita. Human Factors information almost every day

Challenger loss of control in-flight by A380 wake vortex encounter

Interim Report, published May 2017. Bundesstelle für Flugunfalluntersuchung. BFU – German Federal Bureau of Aircraft Accident Investigation


Type of Occurrence: Accident. Date: 7 January 2017. Location: Enroute, above the Arabian Sea. Manufacturer / Model: 1) Bombardier / CL-600-2B16 (604 Variant) 2) Airbus / A380-861. Injuries to Persons: 1) Two severely injured passengers, two passengers and one flight attendant suffered minor injuries 2) None. Damage: 1) Aircraft severely damaged 2) None.

wake 6

Photos: Airbus A380-861 (C) Tim Bowrey. – Jetphotos.net  Bombardier CL-600-2B16 (C) Aktug Ates – Jetphotos.net

Factual Information

During cruise flight above the Arabian Sea, the Indian Ocean, approximately one minute after it had been passed overhead by an Airbus A380 on opposite course, the CL604 was subject to temporary loss of control.

After it had lost approximately 9,000 ft of altitude the pilots regained control of the aircraft and subsequently landed at an alternate aerodrome at Muscat Airport, Oman.

The accident occurred over international waters. Thus the BFU as representative of the State of Registry of the accident aircraft is responsible for the conduct of the investigation. In accordance with international regulations, the air accident investigation authorities of Oman, India, the United Arab Emirates, Canada, USA, and France will assist the BFU in this investigation.

History of the Flight

At 1152 hrs -0652 UTC (All times local, unless otherwise stated) the CL604 had taken off from runway 36 at Malé, Maldive Islands, for a flight to Al-Bateen, United Arab Emirates. Three crew members and six passengers were on board the airplane.

The Flight Data Recorder (FDR) recordings show that the CL604 autopilot had been engaged approximately one minute after take-off. At 0720 UTC the airplane reached cruise level FL340. At 0729 UTC the aircraft entered Indian airspace (Mumbai FIR) at the reporting point BIBGO and had received the clearance to fly to reporting point KITAL via route L894. At approximately 0818 UTC the co-pilot radioed reaching reporting point GOLEM.

At 0655 UTC an Airbus A380-861 (A380) had taken off at Dubai Airport, United Arab Emirates, for a flight to Sydney, Australia. The aircraft flew at FL350 with a southern heading.

The analysis of the flight data of both aircraft showed that at 0838:07 UTC the A380 had passed the CL604 overhead with a vertical distance of 1,000 ft.

At 0838:54 UTC the CL604, with engaged autopilot, began to slightly roll right. At the same time, a counter-rotating aileron deflection was recorded and fluctuation of the vertical acceleration began. In the subsequent approximately 10 seconds the airplane had a right bank angle of 4° to 6°. At 0839:03 UTC the right bank angle began to increase. Within one second the bank angle increased to 42° to the right. At the same time, the aileron deflection to the left increased to 20° and the vertical acceleration increased to 1.6 G. In the following second, vertical acceleration changed to -3.2 G.

At 0839:04 UTC a lateral acceleration of 0.45 G to the right was recorded. The pitch angle changed from about 3° to about 1°, then within one second increased to 9° and decreased again in the following second to -20°. At the same time, the FDR recorded a rudder deflection to the left reaching 11.2° after about two seconds whereas the bank angle changed from 42° right to 31° left.

Between 0839:05 UTC and 0839:10 UTC Indicated Airspeed (in knots) changed from approximately 277 KIAS to 248 KIAS. The N1 of the left engine of 95% began to decrease.

At 0839:07 UTC the validity of IRS parameter is lost, the lateral acceleration reached 0.94 g left, the autopilot disengaged, and a master warning, lasting seven seconds, was recorded.

Between 0839:09 UTC and 0839:41 UTC the FDR recorded a loss of altitude of approximately 8,700 ft. Large control surface deflections and acceleration were recorded. The speed increased and at 0839:31 UTC reached approximately 330 KIAS. At 0839:30 UTC the spoilers extended and 13 seconds later were retracted again. The N1 of the left engine had decreased to approximately 40% when the Interstage Turbine Temperature (ITT) began to increase and nine seconds later had reached 850°. The left engine was shut off.

At about 0856 UTC the Pilot in Command (PIC) informed the air traffic controller in Mumbai of the occurrence, declared the emergency and reported their position, altitude and their intention to fly via KITAL to Oman.

At about 0915 UTC the crew restarted the left engine. Subsequently, the airplane climbed to FL250. At about 0956 UTC the autopilot was re-engaged.

At 1105 UTC the CL604 landed at Muscat Airport.

The A380 continued the flight to Sydney and landed there at 1958 UTC.

The recordings of the Omani air traffic control services show that at about 0920 UTC the neighbouring Indian regional air traffic control Mumbai informed them that the CL604 was at FL230 and would probably pass the reporting point KITAL at 0937 UTC. Mumbai also informed ATC that via a relay station the information had been received that the airplane would divert to Oman. Initially, the reason for the low altitude was given by Mumbai ATC as being due to engine failure. At 0957:50 UTC the airplane was depicted on the Omani ATC radar. At 1014:14 UTC the CL604 reached reporting point KITAL.

Statements of the CL604 Pilots

According to the statement of the CL604 pilots, the PIC was Pilot Flying (PF) and the co-pilot Pilot Non Flying (PNF). The PIC stated that TCAS had drawn his attention to the opposite traffic. He then recognised the aircraft type A380, the airline, and informed the co-pilot. The PIC also stated that the A380 had passed them in opposite direction, slightly to the left and according to TCAS 1,000 ft above. He further stated that a short time later the airplane had been hit by the wake turbulence of the A380. The airplane had shook briefly, then rolled heavily to the left and the autopilot disengaged. Both pilots had actuated the aileron to the right in order to stop the rolling motion. But the airplane had continued to roll to the left thereby completing several rotations. Subsequently both Inertial Reference Systems (IRS), the Flight Management System (FMS), and the attitude indication failed. According to the pilots’ statements at the time of the accident both pilots had fastened their lap belts and in addition the co-pilot had worn his shoulder belts. According to the PIC he had lost his headset during the rolling motion of the airplane. The Quick Reference Handbook (QRH) had flown around the cockpit and was damaged. As a result individual pages had been scattered around the cockpit. The PIC explained since the sky had been blue and the ocean’s surface almost the same colour he had been able to recognize the aircraft’s flight attitude with the help of the clouds. Later both pilots had been able to recover the airplane at FL240 using control inputs on the aileron and later the rudder and slight elevator deflection. Regarding the left engine the PIC stated that he had observed that N1 and N2 had “run apart”. N1 had decreased severely. ITT had increased, reached more than 1,000°C, and the indication flashed red. Subsequently the engine was shut off. Based on the memory items the pilots were able to reactivate the IRS in attitude mode and fly the airplane again towards reporting point KITAL. Then the pilots used the cross bleed of the right engine to restart the left. After the second IRS had been reactivated and position and heading been entered manually into the FMS the autopilot was engaged again. After they had assessed the situation the flight crew decided to fly to Muscat.

Statements of the CL604 Flight Attendant

The flight attendant stated in an interview conducted by the BFU that during take-off and climb she had been seated in the jump seat with the seat belt fastened. She had opened the seat belt while they were passing FL100. At the time of the accident, she had been standing in the middle of the cabin preparing the service. Four of the six passengers had also not been seated. In her recollection, the airplane had turned three times around its longitudinal axis, during which the occupants had been thrown against the ceiling and the seats. Several of the passengers suffered injuries, some of which were bleeding. She herself suffered minor injuries. Using the on-board first aid kit she had attended to the passengers. In the further course of the flight she informed the pilots of the situation in the cabin and reassured the passengers.

Reconstruction of the encounter of the two airplanes

wake 3

wake 4

wake 5

Images: Interim Report Bundesstelle für Flugunfalluntersuchung. BFU – German Federal Bureau of Aircraft Accident Investigation 

Personnel Information

Pilot in Command CL604

The 39-year-old PIC held an Air Transport Pilot’s License (ATPL(A)) of the European Union issued in accordance with Part-FCL. It was first issued by the Luftfahrt-Bundesamt (LBA) and valid until 6 June 2014. The licence listed the ratings as PIC for CL604/605 and the Instrument Rating (IR) valid until 31 March 2017, and for single engine piston land (SEP).

His class 1 medical certificate was last issued on 26 September 2016 and valid until 8 October 2017.

His total flying experience was about 5,334 hours, about 4,564 hours of which were on type.

He had been employed by the operator as a pilot since October 2012.

On the day of the accident, the entire crew had begun their shift at 0500 UTC.

Co-pilot CL604

The 41-year-old co-pilot held an Commercial Pilot’s License (CPL(A)) of the European Union issued in accordance with Part-FCL. It was first issued by the LBA on 31 October 2013. The licence listed the ratings as co-pilot for CL604/605 and the Instrument Rating (IR), valid until 31 October 2017, and for single engine piston land (SEP) and Touring Motor Glider (TMG).

His class 1 medical certificate was last issued on 8 March 2016 and valid until 8 April 2017.

The co-pilot had a total flying experience of about 1,554 hours; of which 912 hours were on type.

Since November 2015 the co-pilot had been employed by the operator.

Meteorological Information

Pre-flight Meteorological Preparation CL604

The BFU was provided with the pre-flight preparation documentation of the CL604

flight crew including the weather data of 6 January 2017 at 2336 UTC.

According to the forecast tropopause was at approximately FL525 at a temperature

of -82°C.

For cruise level FL340 wind with 20 kt from north-west and a temperature of -42°C

were forecast.

The Significant Weather Fixed Time Prognostic Chart for the planned flight did not

contain any warnings of Clear Air Turbulence (CAT) for the area of the Arabian Sea.

Weather at the Time of the Accident

At the time of the accident it was daylight. According to the CL604 pilots’ statements very good Visual Meteorological Conditions (VMC) with blue skies prevailed. The ocean’s surface had been visible. In an estimated altitude of 3,000 to 4,000 ft AMSL the cloud cover had been 1/8 to 2/8. Condensation trails had not been visible.

No significant meteorological information (SIGMET) had been issued for the flight information region Mumbai (VABF).

According to the Digital Access Recorder (DAR) of the A380 the wind at their cruise level at FL350 came from about 315° with about 23 kt. The Static Air Temperature (SAT) was -44°C.

Wreckage and Impact Information

The accident occurred above international waters, the Arabian Sea, approximately 500 NM from any land.

The aircraft manufacturer determined that the airframe structure could not be restored to an airworthy state as it exceeded the airframe certification design load limits during the upset encounter. Therefore the aircraft is considered to be damaged substantially.

During a BFU investigation of the airplane no outer damages on fuselage, wings, and empennage, including control surfaces, were visible. There was no evidence of leakages (oil, fuel).

The inside of the passenger cabin showed damages on the seats and the panelling, as well as traces of blood. The armrests of the four seats in the front, installed in club arrangement, were either deformed or had fractured.

On the left side of the cabin two oxygen masks had fallen from their casings.

wake 1


Photo of the interior of D-AMSC after the upset (Photo: FlightServiceBureau) from The Aviation Herald   

Medical and Pathological Information

According to the operator, four passengers were treated at the hospital in Muscat.

One passenger suffered head injuries and a broken rib; another passenger had fractured a vertebra. The two passengers and the flight attendant, who had sustained minor injuries, suffered bruising and a fractured nose, respectively.

The two other passengers and the pilots remained unharmed.

Additional Information

Safety Case for Wake Vortex Encounter Risk due to the A380-800

An ad hoc Steering Group (SG) and a technical Work Group, comprising representatives from Joint Aviation Authorities (JAA), Eurocontrol, Federal Aviation Administration (FAA), Airbus and Det Norske Veritas (DNV), was set up in 2003 to specify safety requirements to ensure Wake Vortex Encounter (WVE) risk from the Airbus A380 will be acceptable. A safety case (A380 SG, 2006a) and supporting documentation has been produced.

Among others the following recommendations have been made:

wake 2Investigator in charge: Jens Friedemann

Excerpted from Bundesstelle für Flugunfalluntersuchung – German Federal Bureau of Aircraft Accident Investigation Interim Report


On Mar 18th,  2017 The Aviation Herald received a draft of an EASA safety information bulletin regarding this accident, stating:

With the increase of the overall volume of air traffic and enhanced navigation precision, wake turbulence encounters in the en-route phase of flight above 10 000 feet (ft) mean sea level (MSL) have progressively become more frequent in the last few years.

The aim of this SIB is to enhance the awareness of pilots and air traffic controllers of the risks associated with wake turbulence encounter in the en-route phase of flight and provide recommendations for the purpose of mitigating the associated risks.

The draft reasons:

The basic effects of wake turbulence encounter on the following aeroplane are induced roll, vertical acceleration (can be negative) and loss or gain of altitude. The greatest danger is an induced roll that can lead to a loss of control and possible injuries to cabin crew and passengers. The vortices are also most hazardous to the following aircraft during the take-off, initial climb, final approach and landing.

However, en-route, the vortices evolve in altitudes at which the rate of decay leads to a typical persistence of 2-3 minutes, with a sink rate of 2-3 metres per second. Wakes will also be transported by the wind.

Considering the high operating airspeeds in cruise, the wake can be encountered up to 25 nautical miles (NM) behind the generating aeroplane, with the most significant encounters reported within a distance of 15 NM. This is larger than in approach or departure phases of flight.

The encounters are mostly reported by pilots as sudden and unexpected events. The awareness of hazardous traffic configuration and risk factors is therefore of particular importance to anticipate, avoid and manage possible wake encounters.

The draft issues following recommendations:

As precautionary measures, operators and pilots should be aware that:

– As foreseen in Reg. 965/2012 AMC1 to CAT.OP.MPA.170, the announcement to passengers should include an invitation to keep their seat belts fastened, even when the seat belt sign is off unless moving around the cabin. This minimises the risk of passenger injury in case of a turbulence encounter en-route (wake or atmospheric).

– As indicated in ICAO PANS-ATM, for aeroplanes in the heavy wake turbulence category or for Airbus A380-800, the word “HEAVY” or “SUPER”, respectively, shall be included immediately after the aeroplane call sign in the initial radiotelephony contact between such aeroplanes and ATS units.

– When possible, contrails should be used to visualise wakes and estimate if their flight path brings them across or in close proximity.

– When flying below the tropopause altitude, the likelihood of wake encounter increases. The tropopause altitude varies (between days, between locations).

– Upwind lateral offset should be used if the risk of a wake encounter is suspected.

– Timely selecting seat belt signs to ‘ON’ and instruct cabin crew to secure themselves constitute precautionary measures in case of likely wake encounters.

In the case of a wake encounter, pilots should:

– Be aware that it has been demonstrated during flight tests that if the pilot reacts to the first roll motion when in the core of the vortex, the roll motion could be amplified by this initial piloting action. The result can be a final bank angle greater than if the pilot would not have moved the controls.

– Be aware that in-flight incidents have demonstrated that pilot inputs may exacerbate the unusual attitude condition with rapid roll control reversals carried out in an “out of phase” manner.

– Be aware that if the autopilot is engaged, intentional disconnection can complicate the scenario, and the autopilot will facilitate the recovery.

– Avoid large rudder deflections that can create important lateral accelerations, which could then generate very large forces on the vertical stabiliser that may exceed the structural resistance. Although some recent aircraft types are protected by fly-by-wire systems, use of the rudder does not reduce the severity of the encounter nor does it improve the ease of recovery.

– Make use of specific guidance available through AOM for their specific type(s)/fleet.

ATS providers and air traffic controllers should:

Enhance their awareness about en-route wake turbulence risk, key factors and possible mitigations, based on the information provided in this document and other relevant material. This could be achieved through flyers, e-learning, and refresher training module.

Possible risk mitigations may consist of:

– Make use of the wake turbulence category (WTC) indication in the surveillance label and/or the flight progress strip (whether electronic or paper), and observe closely separated aeroplanes that are at the opposite extremes of the WTC spectrum;

– As the best practice, provide traffic information, advising “CAUTION WAKE TURBULENCE”, when you identify that a ‘HEAVY’ or ‘SUPER HEAVY’ wake category traffic is climbing or descending within 15 NM of another following traffic;

– Manage en-route traffic crossings such as, when possible while preserving safe tactical management of overall traffic in the sector, avoiding to instruct climb or descend to ‘HEAVY’ or ‘SUPER HEAVY’ traffic within 15 NM distance from another following traffic;

– If at all possible, avoid vectoring an aeroplane (particularly if it is LIGHT or MEDIUM category) through the wake of a HEAVY or SUPER HEAVY aeroplane where wake turbulence may exist.


  1. Bundesstelle für Flugunfalluntersuchung – German Federal Bureau of Aircraft Accident Investigation Interim Report. State File Number: BFU17-0024-2X. Published: May 2017
  2. EASA  Safety Publications Tool. Safety Information Bulletins
  3. The Aviation Herald


minime2By Laura Victoria Duque Arrubla, a medical doctor with postgraduate studies in Aviation Medicine, Human Factors and Aviation Safety. In the aviation field since 1988, Human Factors instructor since 1994. Follow me on facebook Living Safely with Human Error and twitter@dralaurita. Human Factors information almost every day

Let’s go around

Runway excursion has been the most frequent category of aviation accidents for many years now, and according to Flight Safety Foundation studies 83% of these accidents could have been prevented with a go-around. Moreover, 54% percent of all accidents could be prevented by going around, meaning the go-around is perhaps one of the most critical strategies in preventing aviation accidents.

Go around

Photo from Radko Našinec video Crosswind landing results in almost crash / Boeing 737 hard touchdown + go around, Prague (LKPR)

However, compliance with go-around policies is extremely poor among the industry, only 3% of unstable approaches result in go-around policy compliance.

On the other hand, going around is not risk-free. During the go-around, there is an increased risk of loss of control-LOC events, as we have seen in several major accidents in the last two decades, but there are risks of controlled flight into terrain-CFIT and midair collision-MAC too.

As Flight Safety Foundation states, in the Go-Around Decision-Making and Execution Project final report, these risks associated with conducting a go-around could be triggered by:

  • Ineffective initiation of a go-around, which can lead to LOC;
  • Failure to maintain control during a go-around, which can lead to LOC, including abnormal contact with the runway or to CFIT;
  • Failure to fly the required track, which can lead to CFIT or MAC;
  • Failure to maintain traffic separation, which can lead to MAC; and,
  • Generation of wake turbulence, which may create a hazard for another aircraft that can lead to LOC.

Therefore, the go-around decision should balance the risk associated with the continuing to land with the risk associated with the go-around itself.

Besides the operational, procedural and technical aspects, there are several Human Factors issues associated with a go-around decision and execution. These Human Factors lessons have been learned from incidents and accidents associated with going around and have been studied further and thorough by several organizations around the world. Let’s review some of them:

  • Although a go-around is considered as a normal procedure, a BEA study showed that a go-around does not often occur during operations, could be complex in terms of workload and is one of the maneuvers that are poorly represented by simulators. For all these reasons, “in practice, the go-around procedure is not a normal procedure but a specific one”. (Bureau d’Enquêtes et d’Analyses (BEA) pour la Sécurité de l’Aviation civile. Study of Airplane State Awareness during Go-Around. August 2013)
  • A go-around is a normal but no routinely executed procedure for most commercial pilots. A long-haul commercial pilot may conduct one go-around every two to three years whereas a short-haul commercial pilot may conduct a go-around once or twice a year. (Flight Safety Foundation Go-Around Decision-Making and Execution Project final report)
  • Is poorly represented by full flight simulators, due to no realistic ATC environment and the SIM inability to represent the physiological sensations associated with a go-around
  • Produces a disruption that often is unexpected by the flight crew, therefore, can induce the startle effect
  • It breaks the continuity of tasks being performed and suddenly demands a new set of actions with discontinuous tasks and disrupted rhythms of execution
  • It comprehends many tasks of different nature which must be performed in a limited period of time
  • The maneuvers required are varied and must be performed rapidly
  • The parameters that must be controlled are numerous and rapidly changing
  • Besides to monitor attitude, thrust, flight path, aircraft configuration and pitch trim, pilots have to monitor the autopilot, the flight director, the autothrottle, their modes, cross check one to each other and the airplane to be sure they themselves and the plane are doing the right thing
  • All of the above could produce an information overload, high workload and stress especially when the startle effect is also present
  • This sudden high workload is higher for the Pilot Monitoring-PM than that for the Pilot Flying-PF. PMs have to deal with the readbacks of ATC instructions, the callouts, the monitoring of the PF’s flight control, the verification of the pitch attitude and the verification of flight mode annunciator (FMA) modes. This high workload often prevents the PM from fulfilling the task of monitoring the PF
  • As stress can reduce the ability to execute complex actions, the higher the level of stress the higher the performance can be compromised
  • Stress also could lead to attention channelling which produces excessive focusing on some task while neglecting others
  • “Automatic systems could add to the problems because their initial engagement modes are different from those expected for the go-around … and when they are neither called out nor checked, this leads the aeroplane to follow an unwanted flight path” (BEA Study on Aeroplane State Awareness During Go-Around)
  • All these, the startle effect, the time pressure, the automation issues, the cognitive overload, the potentially overwhelming situation could produce a degradation of CRM skills
  • Spatial disorientation could be an aggravating factor
  • Somatogravic illusions induced by the linear acceleration produced by the full thrust, maybe in a relatively light at the end of a flight plane, can cause the PF to reduce the pitch angle. This may induce loss of control during the go-around. Full flight simulators lack the ability to accurately represent a somatogravic illusion
  • Low relevant experience of one or both pilots can affect the effectiveness of monitoring during go-around
  • Complex arrivals, departures and go-around procedures increment the workload for the flight crews and could be another aggravating factor
  • In the same way, the intervention of ATC with too much information in a radio transmission can lead to pilot confusion
  • Changes to go-around instructions increase the already high workload for pilots. Same could happen with late provision of go-around instructions
  • Sometimes ATC instructions are not compatible with aircraft performance
  • Bringing out unpublished go-around tactical instructions can place high demands on pilots

The decision to go around is perhaps the one with most impact has in aviation accident reduction. How can the risk associated be mitigated and how compliance with go-around policies can be improved?

Answering that question is obviously far beyond the scope of this article. There are now numerous studies on the subject with the results of research, findings and analysis, with proposed strategies and recommendations for the industry, the regulators, the operators, the flight crews and the air traffic services providers. Nonetheless, several factors can be stressed and highlighted here:

  1. “Pilots and their employers should understand that one of the many reasons that violating approach minimums is unacceptable is because evidence indicates that if a go-around then becomes necessary, the chances of a safe transition to the go-around are reduced.” (Flight Safety Foundation Go-Around Decision-Making and Execution Project final report). So NOT violate approach minimums
  2. “The lack of decision is the leading risk factor. In other words, if you’d made your decision earlier in the process, you’d probably able to execute the go-around better than forced into it by having an unstabilized approach, then, at the very last second, deciding have to go around,” (Dave Carbaugh, presentation to Flight Safety Foundation’s 67th annual International Air Safety Summit, 2014) So DO NOT delay the decision, it could prevent the situation to escalate.
  3. “Ensure that go-around training integrates instruction explaining the methodology for monitoring primary flight parameters, in particular, pitch, thrust, then speed.” (BEA Study on Aeroplane State Awareness During Go-Around) It seems obvious, but apparently, it has not been.
  4. “Ensure that go-around training and awareness appropriately reflect different go-around execution risk scenarios” (Flight Safety Foundation Go-Around Decision-Making and Execution Project final report) 
  5. Enhance training “… including realistic detailed training scenarios based on current technology and risks” (BEA Study on Aeroplane State Awareness During Go-Around)
  1. “Review go-around policy, procedures and documentation to maximize their effectiveness, clarity and understanding.” (Flight Safety Foundation Go-Around Decision-Making and Execution Project final report)
  2. “Study the additional technical and regulatory means required to mitigate the shortcomings of CRM in high workload and/or unusual conditions” (BEA Study on Aeroplane State Awareness During Go-Around)
  3. “Air traffic controllers, except where necessary for safety reasons, do not give instructions that are in contradiction with the published missed-approach procedure; and that, when necessary, the instructions are announced to crews as early as possible during the approach.” (BEA Study on Aeroplane State Awareness During Go-Around)

As Captain Dave Carbaugh, stated in his presentation to Flight Safety Foundation’s 67th annual International Air Safety Summit (IASS) in Abu Dhabi, United Arab Emirates, in November 2014 “30 years ago we had low thrust-to-weight ratios, so the airplane just didn’t climb very quickly. We had less traffic density, so there wasn’t anybody in front of you and … and a lot of times, the go-round was non-complex. ‘Fly runway heading to 4,000. It was basically easy. But those days are gone…”


  1. Blajev, Tzvetomir, Curtis, William. Go-Around Decision-Making and Execution Project. Final report to Flight Safety Foundation. March 2017
  2. Bureau d’Enquêtes et d’Analyses (BEA) pour la Sécurité de l’Aviation civile. Study of Airplane State Awareness during Go-Around. August 2013
  3. Rosenkrans, Wayne. Go-Around Risks. AeroSafety World April 2015. https://flightsafety.org/asw-article/go-around-risks/


  1. MyCargo B744 fatal accident at Kyrgyz Republic, Jan 16th, 2017. Preliminary Report
  2. Going around with no thrust. Emirates B773 accident at Dubai on August 3rd, 2016, interim report
  3. The Head-Up Illusion: do you remember it?
  4. Armavia A320 crash during go-around at night in poor meteorological conditions
  5. Tatarstan B735 crash during go-around at night. Learning from the recent past
  6. Going around with all engines operating
  7. Speaking of going around
  8. Loss of flight crew airplane state awareness 


minime2By Laura Victoria Duque Arrubla, a medical doctor with postgraduate studies in Aviation Medicine, Human Factors and Aviation Safety. In the aviation field since 1988, Human Factors instructor since 1994. Follow me on facebook Living Safely with Human Error and twitter@dralaurita. Human Factors information almost every day

Women and aviation safety

woman in comandPhoto: Elizabeth L. Remba Gardner, Class 43-W-6 WASP (Women Airforce Service Pilot) at the controls of a Martin B-26 ‘Marauder’ medium bomber. Harlingen Army Air Field, Texas. 1943

In 2013, the United States Army published a study on the role of women in combat, which shows how, in the years 2002 to 2013 women were involved in fewer aviation accidents than male crews, only 3% of accidents. With women accounting for approximately 10% of flyers, the evidence could suggest, according to the study, that women can operate an aircraft more safely. Concerning only the AH64-APACHE, 100% of the accidents in the garrison and in the theater of operations, involved male crews, which would suggest, according to the author of the study, that female attack pilots could be even More secure in the performance of flight activities. (1)

Is this a generalized tendency or is it specific to the population studied?

In 1986 a study was published in which, when analyzing by gender the NTSB files of all general aviation accidents occurring between 1972 and 1981, a higher accident rate was found in men than in women and a higher rate proportion of deaths or serious injuries in the accidents of male pilots in those of female pilots. These differences were found in all variables analyzed: type of license, age, total flight time, flight time in aircraft type, operating phase, flight category, specific cause and causal factors. (2)

On the other hand, a study published in 1996 analyzed if there were differences in the accident rate between male and female airline pilots in the United States, based on the analysis of information obtained from the FAA on accidents between 1986 and 1992. Initially, it was found that in general women hired by major airlines had significantly higher accident rates than men. However, the study emphasizes that on average women were much younger and less experienced than male pilots, whereby male pilot accidents were adjusted for age, experience (total flight time), risk exposure (hours flown in the previous 6 months) and airline (major airline vs non-major airline), using logistic regression. After adjusting the variables it was found that there is no difference in the rate of accidents of female and men pilots, which suggests that neither women nor men are a safer group than the other. (3) (NOTE: Logistic regression is a special type of regression that is used to express and predict the probability between two groups that an event occurs, given a series of independent variables)

The same author published in 1997 a similar study analyzing this time the incident rate in a population of 70,164 airline pilots. The regression analysis again indicated that there are no significant differences in performance between female and male airline pilots. (4)

On the same track, in 1998 Caldwell & Le Duc found that in combat pilots, gender did not produce significant operational effects on flight or mood performance and the ability to cope with stressors associated with combat. (5)

However, for general aviation, the causes of air accidents do seem to be related to gender, as shown in a study published in 2001 in which the fatal and non-fatal per capita accident rate was higher in men than in women (3,20), which would confirm the findings of the 1986 study previously mentioned (6)

Finally, in 2002, a study was conducted at Embry Riddle Aeronautical University which found that there are no aspects of a captain’s competence that are related to gender. The author of the study even poses the possibility that this lack of difference in performance is due to the personality traits present in pilots, whether men or women. The above, based on a study that showed through psychological tests that the personality profile of female drivers has little resemblance to the profile of adult women in the US, followed by a high resemblance to the profile of adult men in the US and with the closest resemblance to the male pilot profile. This personality similarity, says the study’s author, may have eliminated differences in skills that have been apparent in other work groups that do not require, attract, and select such a specific personality type.

female 1

Photo: Capt Kerstin Felser

Based on the above we could conclude that although for some types of aviation it is true that pilots show better performance indices than their male counterparts, this difference does not occur in commercial aviation. Better yet, there is no reason to think that there is any difference in performance capability between airline pilots of both genders.

Therefore, there is no reason to prefer one over another in the selection processes to enter the airline aviation that is the one that occupies us, and that for obvious reasons, generates more attention in the public, the media and the Governments.

So why are not there more female pilots? I do not have the answer to that question, but I can tell you that for reasons of air safety, it is not.

Although there is still controversy about whether or not there are differences in gender-related cognitive skills, any variation that exists has very little relevance in air operations. (Caldwell & LeDuc, 1998).


  1. Peña-Collazo, Seneca. Women in Combat Arms: A Study of the Global War on Terror. A Monograph.S. Army, School of Advanced Military Studies. United States Army Command and General Staff College. Fort Leavenworth, Kansas. January 2013. Page 47
  2. Vail GJ, Ekman LG, Pilot-error accidents: male vs female. Applied Ergonomics. 1986 Dec;17(4):297-303.
  3. L. McFadden, Comparing pilot-error accident rates of male and female airline pilots. Omega Volume 24, Issue 4, August 1996, Pages 443-450.
  4. Kathleen L. McFadden, Predicting pilot-error incidents of US airline pilots using logistic regression. Applied Ergonomics Volume 28, Issue 3, June 1997, Pages 209-212
  5. Caldwell JA Jr, LeDuc PA, Gender influences on performance, mood and recovery sleep in fatigued aviators. 1998 Dec;41(12):1757-70.
  6. Baker SP, Lamb MW, Grabowski JG, Rebok G, Li G, Characteristics of general aviation crashes involving mature male and female pilots. Aviation Space and Environmental Medicine. 2001 May;72(5):447-52
  7. Paulsen, Marianne, Perception of Competence in Male and Female Pilots: Between Group Differences. Embry-Riddle Aeronautical University – Daytona Beach. Spring 2002


This is a translation of the article Mujeres y Seguridad Aérea, which was originally published on SEPLA-Sindicato Español de Pilotos de Linea Aérea (Spanish Airline Pilots Union) website.


minime2By Laura Victoria Duque Arrubla, a medical doctor with postgraduate studies in Aviation Medicine, Human Factors and Aviation Safety. In the aviation field since 1988, Human Factors instructor since 1994. Follow me on facebook Living Safely with Human Error and twitter@dralaurita. Human Factors information almost every day