B738, Georgetown Guyana, 2011
From SKYbrary Wiki
|On 30 July 2011, a Boeing 737-800 overran the wet landing runway at Georgetown after a night non-precision approach, exited the airport perimeter and descended down an earth embankment. There were no fatalities but the aircraft sustained substantial damage and was subsequently declared a hull loss. The Investigation attributed the overrun to a touchdown almost two thirds of the way down the runway and failure to utilise the aircraft’s full deceleration capability. Loss of situational awareness and indecision as to the advisability of a go-around after a late touchdown became inevitable was also cited as contributory to the outcome.|
|Actual or Potential
|Human Factors, Runway Excursion|
|Domicile||Trinidad and Tobago|
|Type of Flight||Public Transport (Passenger)|
|Origin||Piarco International Airport|
|Intended Destination||Georgetown/Cheddi Jagan International|
|Take off Commenced||Yes|
|Location - Airport|
|Airport||Georgetown/Cheddi Jagan International|
|Tag(s)||Deficient Crew Knowledge-systems,|
Deficient Crew Knowledge-handling,
Copilot less than 500 hours on Type
|Tag(s)||Inappropriate crew response - skills deficiency,|
Procedural non compliance,
Ineffective Monitoring - PIC as PF
|Tag(s)||Overrun on Landing,|
Incorrect Aircraft Configuration,
Ineffective Use of Retardation Methods,
Continued Landing Roll
|Tag(s)||Evacuation slides deployed|
Evacuation on Pax Initiative
|Damage or injury||Yes|
|Aircraft damage||Hull loss|
|Causal Factor Group(s)|
On 30 July 2011, a Boeing 737-800 (9Y-PBM) being operated by Caribbean Airlines on a scheduled international passenger flight from Port of Spain Trinidad to Georgetown as BW 523 overran the landing runway at destination after completing a night RNAV(GPS) approach in VMC. It descended down a 20 foot high earth embankment and stopped only after it had exited the airport perimeter and sustained such extensive damage that it was subsequently declared a hull loss. All 163 occupants successfully evacuated the wreckage but one serious injury and an unspecified but multiple number of minor injuries were sustained as a result of either the impact or the evacuation.
An Investigation was carried out by the Guyana Civil Aviation Authority (GCAA) with technical assistance from the NTSB. Data relevant to the Investigation was obtained from both the FDR and CVR fitted to the aircraft.
It was noted that the 52 year-old Captain, who had been PF for the accident flight, had accumulated a total of 16,600 flying hours of which 5,000 hours were on type. He had previously flown the MD83 and had been a pilot with BWI Airlines and its successor Caribbean Airlines for 25 years. The 23 year-old First Officer had 1,400 total flying hours of which 350 hours were on type. He had been with Caribbean Airlines for three years and had previously flown as a First Officer on the DHC8.
It was established that the flight had begun in New York and had made a transit stop at Port of Spain, Trinidad where a complete crew change also occurred prior to departing on the one hour flight to Georgetown. FDR data indicated that the flight had been routine apart from a deviation around a thunderstorm cell as the aircraft began descent and positioning for an approach to runway 06 to which a clearance for an RNAV(GPS) approach was given. With about 5 nm to go to touchdown, flap 30 was selected and during the stabilised approach, the AP was disconnected when passing around 1,000 feet. After a further 45 seconds, the aircraft crossed the runway threshold at speed close to the applicable Vref of 149 KIAS. However, it then floated with excess thrust maintained and main gear touchdown on the wet surface did not occur until the aircraft was approximately 1,430 metres along the 2,270 metre-long runway.
A few seconds before touchdown occurred, the CVR recorded the Captain calling a go-around and the First Officer acknowledging it but when, three seconds after main gear touchdown, the First Officer called “Max?”, the Captain quickly responded that he “had it” and thrust was not required. The First Officer was then recorded immediately calling autobrake disarmed and manual braking as automatic spoiler deployment followed. Thrust reversers were selected without delay but with just 610 metres to go until the end of the runway and only to a position between idle and detent 2. Manual brake pressure was only slowly increased and did not reach the maximum achievable pressure of 3000 psi until there was only 75 metres to go until the end of the runway. The prolonged float and delayed touchdown were attributed to the maintenance of excessive thrust when over the first quarter of the runway.
As it continued down the runway, the aircraft began to slowly but progressively deviate left of the centreline. It subsequently departed the end of the runway, descended down a 20 foot high slope and broke through the chain link perimeter fence before it crossed a dirt track and came to a stop some 40 metres beyond the end of the runway. At the time of the landing, the wind was light and variable and was assessed to have had minimal effect on the controllability of the aircraft.
Once the aircraft had come to a stop, it was reported that the interior of the aircraft was in almost complete darkness. The Purser reported that as the landing continued, she had realised that something was wrong and had shouted commands to the passengers to “bend over and get your heads down” three times. There was no PA from the flight deck and the evacuation was initiated by passengers, one of whom opened the R1 door and the slide inflated. Several other passengers followed him but L1 door was jammed shut by a Galley Unit. The Purser followed the passengers out of the R1 door and once outside used her flashlight to guide passengers to an assembly area at the front of the aircraft. It was not possible to do a headcount as the large group in the assembly area was a mixture of passengers and emergency services personnel and it was not possible to distinguish between them in the darkness. By this time she reported that “several vehicles had arrived on the scene” but as none of them were ambulances, a seriously injured passenger she had located was “placed in the back of a pick-up truck”. In the rear cabin, efforts to open the L2 door were unsuccessful and it was suspected that this was attributable to lack of crew familiarity with the opening of exits when an aircraft is in an unusual attitude rather than any jamming or malfunction of the door. It then took two people to open the R2 door which was eventually opened with difficulty and only “five to ten” passengers were able to use this exit before its slide deflated. One of the cabin crew then jumped to the ground and got a passenger to help hold the slide out like a chute and about another ten passengers were able to use this door. The over-wing emergency exits functioned properly and were used by a majority of the passengers.
The flight crew stated that due to the unexpected nature of the occurrence, they had been in no position to “alert the cabin crew or passengers about the impending disaster (and) had no opportunity to carry out any emergency drills”. The First Officer reported that after the impact he had removed his harness and got out of his seat. He used his flashlight to illuminate the flight deck which was completely dark. The door into the passenger cabin was ajar but as with the two flight deck side windows, it was jammed and exit into the cabin was also obstructed by the forward toilet door which was jammed open. Eventually, the First Officer managed to crawl into the cabin and RFFS personnel arrived and helped get the injured Captain out of the flight deck.
The Investigation identified and reviewed various elements in the circumstances in which the accident occurred. These included the following:
- Wet Runway Landings - It was found that the Operator had issued explicit guidance for wet runway landings and that in such circumstances, flaps 40° should be used when landing on runways less than an 8,000 feet (2,438 metre) LDA. It was also reported that pilot recurrent training included go-arounds from below minima (baulked landings). However, the accident Captain “could not recollect the existence of these wet runway procedures or receiving any training in a go round after touchdown” which was taken as an indication that there might have been deficiencies in his recurrent training.
- CRM on the Flight Deck - Whilst CVR data were found to show during most of the flight there had been good coordination between the two pilots with the Captain carrying out his duties professionally and all briefings completed in accordance with applicable Standard Operating Procedures (SOPs), a lack of CRM became apparent as the runway threshold was crossed. The data available were considered to suggest that a temporary loss of situational awareness had affected both pilots with neither being aware of the position of the aircraft in relation to the remaining runway. It was also concluded that there was doubt as to whether either pilot had appreciated that the aircraft was floating as a result of the almost 60% N1 thrust setting maintained over the runway by the Captain. There was certainly “no indication from the CVR that the First Officer initiated any action that would have helped to alleviate the accident results”. It was particularly noted that the decision not to go around when the call had already been made and acknowledged was contrary to the Operator’s Operations Manual Part A which explicitly rules out changing one’s mind after declaring a go-around.
- Absence of a RESA - Runway 06 at Georgetown has no RESA yet a RESA is a Standard required by ICAO Annex 14 and it was considered that the availability of one may have reduced the severity of the accident.
- Limited lighting for approach and landing - Runway 06 at Georgetown has no approach lights, centreline lighting, or REILs and nothing from the fight crew interviews indicated that either pilot had perceived any difficulty with runway lighting or locating the runway. However, it was considered that this lack of lighting combined with the need to transition from an approach over dark jungle terrain to a lighted area may have been conducive to black hole illusion. Depth perception could have also have been affected by the slope of the runway and the fact that it was wet and would have been reflecting light.
- PAPI beam below with instrument approach vertical profile - Although runway 06 is equipped with PAPIs, their beam projection was found to be 2.39º which did not coincide with the 3° vertical profile defined for the RNAV (GPS) procedure flown by the accident aircraft. It was considered that this shallow approach slope could have exacerbated the previously-identified risk of visual illusion due to the lack of approach lighting and the wet runway. In turn, this may have accounted for the excessive thrust maintained by the Captain when over the runway which contributed to the aircraft floating down the runway.
- Absence of ‘Distance to go’ markers - The lighting and marking of the runway met ICAO standards but at the time of the accident there were no distance-to-go markings. Whilst these are not an ICAO requirement, it was considered that had they been present, situational awareness of the pilots as to their position in relation to the end of the runway might have been enhanced.
- Airport Categorisation - It was noted that whilst Caribbean Airlines has established special categories of airports, Georgetown Guyana is not one of them as it is considered by the airline to be similar to and typical of many island approaches over water.
The Probable Cause of the accident was formally documented as having been that “the aircraft touched down approximately 4700 feet beyond the runway threshold, some 2700 feet from the end of the runway, as a result of the Captain maintaining excess power during the flare, and upon touching down, failure to utilise the aircraft’s full deceleration capability, resulted in the aircraft overrunning the remaining runway and fracturing the fuselage”.
Three Contributory Factors relating to the performance of the flight crew were identified as follows:
- Indecision as to the execution of a go-around.
- Failure to execute a go-around after the aircraft floated some distance down the
- Diminished situational awareness.
A total of 23 Safety Recommendations were made as a result of the Investigation as follows:
- that Caribbean Airlines should ensure that the need to confirm with and operate within Standard Operating Procedures is emphasised during training.
- that Caribbean Airlines pilots must be trained in accordance with the landing techniques detailed in the Boeing Operations Manual.
- that Caribbean Airlines must train its pilots on the appropriate use of maximum reverse thrust, spoilers and maximum manual braking.
- that Caribbean Airlines must emphasise the importance of decision making during initial and recurrent pilot training.
- that Caribbean Airlines CRM training must be reviewed to address the roles of the pilot flying and the pilot monitoring especially in relation to the call for “GO AROUND” when it is observed that an aircraft does not touchdown in the touchdown area. The Operations Manual Part A, clearly outlines when a “go around” call must be made and although not applicable to this accident, the Company should consider emphasising that either flight crew member can make the “go around” call and the response to this must be immediate.
- that Caribbean Airlines should conduct an analysis with their pilots to determine the extent of their understanding of go-around from flare to touchdown and when a go-around is mandatory.
- that Caribbean Airlines must reinforce training with regard to recommended flap setting for wet runways.
- that Caribbean Airlines must develop procedures for flight crew to check charts for landings on wet runways.
- that Caribbean Airlines must provide proactive training in TEM and Visual Illusions Awareness.
- that Caribbean Airlines should review the guidance provided by ICAO in Annex 6 to the ICAO Convention on Fatigue Risk Management Systems.
- that Caribbean Airlines training for flight crew in the use of the (flight deck door) blow out panels to facilitate quick escape must be developed.
- that Caribbean Airlines should review cabin crew training to ensure awareness of the correct procedures for use of emergency exits when aircraft are in an unusual attitude.
- that Caribbean Airlines should consider placement of additional information on the (Passenger) Safety Briefing Card to include directions for the safest and easiest way to get off the aircraft wings and what to do after getting out of the aircraft. This should include a warning not to touch the engines cowlings in order to avoid burns.
- that the Trinidad and Tobago CAA should carry out a full review of the Caribbean Airlines flight crew training program to ensure that it is in keeping with the Boeing program. Emphasis must be put on decision making, situational awareness and CRM and the Company will also benefit by placing greater emphasis on TEM.
- that the Trinidad and Tobago CAA must, when reviewing Caribbean Airlines flight crew training programmes, ensure that the required flight and ground training is carried out and is properly documented.
- that the Trinidad and Tobago CAA should review Caribbean Airlines cabin crew training to ensure it is in keeping with requirements of the Training Manual.
- that the Trinidad and Tobago CAA should undertake a general review of Caribbean Airlines record keeping to ensure that documents are properly completed including sign off and dating.
- that the Operator of Georgetown Airport should include the development of a RESA in accordance with ICAO Annex 14 in the planned extension of the runway and the installation of an Engineered Materials Arresting System (EMAS) should be considered.
- that the Operator of Georgetown Airport should ensure that Approach, Touchdown Zone and Centreline lighting should be included as part of the runway extension plans.
- that the Operator of Georgetown Airport should provide an easily identifiable area that is (available in the event of an aircraft accident or incident) to provide comfort to passengers who may be in distress and their relatives.
- that the Operator of Georgetown Airport should arrange for an engineering assessment of the runway slope and provide notification for publication of same in the Guyana AIP.
- that the Operator of Georgetown Airport must carry out an investigation to determine why there was no reaction (to the accident) from the Airport Security Service and implement training exercises to familiarise the service with its responsibility in keeping with the Airport Emergency Manual.
- that the Guyana Civil Aviation Authority (GCAA) should consider the need for a Meteorological Officer to be stationed at a strategic location on the airport to provide local weather information.
Safety Action taken by the Georgetown Airport Operator prior to the completion of the Investigation was noted as having included an adjustment to the PAPI beam projection on runway 06 to match the vertical profile of the RNAV-GPS approach and the installation of distance-to-go signboards at 1,000 foot intervals beginning at 4,000 feet in both directions of the runway.
The Final Report was completed in September 2014, subsequently published and eventually released online.