F50, vicinity Nairobi Kenya, 2014
From SKYbrary Wiki
|On 2 July 2014, a Fokker 50 fully loaded - and probably overloaded - with a cargo of qat crashed into a building and was destroyed soon after its night departure from Nairobi after failing to climb due to a left engine malfunction which was evident well before V1. The Investigation attributed the accident to the failure of the crew to reject the takeoff after obvious malfunction of the left engine soon after they had set takeoff power which triggered a repeated level 3 Master Warning that required an automatic initiation of a rejected takeoff.|
|Actual or Potential
|Airworthiness, Fire Smoke and Fumes, Human Factors, Loss of Control|
|Type of Flight||Public Transport (Cargo)|
|Origin||Jomo Kenyatta International Airport|
|Intended Destination||Mogadishu International Airport|
|Take off Commenced||Yes|
|Flight Phase||Take Off|
|Location - Airport|
|Airport vicinity||Jomo Kenyatta International Airport|
|Tag(s)||Inadequate Airworthiness Procedures,|
Deficient Crew Knowledge-systems,
Deficient Pilot Knowledge
|Tag(s)||Post Crash Fire|
|Tag(s)||Inappropriate crew response - skills deficiency,|
Inappropriate crew response (technical fault),
Plan Continuation Bias,
Procedural non compliance
|Tag(s)||Loss of Engine Power,|
|System(s)||Engine - General|
|Contributor(s)||Inadequate Maintenance Schedule|
|Damage or injury||Yes|
|Aircraft damage||Hull loss|
|Fatalities||Most or all occupants (4)|
|Causal Factor Group(s)|
On 2 July 2014, a Fokker 50 (5Y-CET) being operated by Skyward International on a non-scheduled cargo flight from Nairobi to Mogadishu crashed into a building shortly after takeoff from Nairobi in night VMC and was destroyed by impact forces and a post crash fire. Both pilots and two other unspecified crew members also on board were killed.
An Investigation was carried out by the Aircraft Accident Investigation Department of the Kenyan Government Department of Transport. Relevant data were downloaded from the aircraft FDR and CVR and was also available from the airport surface movement radar at a resolution greater than that available to the runway controller in real time.
It was noted that the 54 year-old Captain, who was PF for the flight, had a total of 14,531 flying hours experience which included 6,821 hours on type in command. He had flown about 77 hours in the accident aircraft in the 60 days prior to the accident and had previous experience on DHC8, F27 and LET 410 aircraft. The First Officer had a total of 823 hours which included 513 hours on type.
It was established that the aircraft involved had been positioned from its home base at nearby Nairobi Wilson Airport the previous day with the accident flight Captain and different First Officer. On arrival, it was loaded with a cargo of qat in preparation for the accident flight. The accident crew consisted of the Capt and FO in addition to a maintenance engineer and loadmaster. According to air traffic control, the flight plan was filed for duration of 2.5 hours at an altitude of 19,000 feet along with 5 hours of fuel aboard.
FDR data showed that 20 minutes after engine start, the flight crew made initial contact with TWR and a takeoff clearance for runway 06 was issued a couple of minutes later. Takeoff power was set with the left engine stabilising at 78% torque and 100% propeller rpm and the right engine stabilising at 80% torque and 99% propeller rpm. The Captain was then recorded saying “power is set” followed by “the left autofeather is off” and the First Officer calls “780”, a possible reference to both engine ITTs.
About 16 seconds after takeoff power had been set, the first of a series of triple chime level 3 alerts occurred whilst at about the same time, FDR data recorded a drop in left engine fuel flow and ITT - although the engine torque recorded for both engines remained at 80%. These triple chime alerts then continued at one second intervals for the remainder of the flight but there was no recorded verbal response to them from either pilot. About 24 seconds after takeoff power had been set with the ground speed around 30 knots, the left engine torque rose quickly to the maximum figure which can be recorded - 120% - whilst at the same time, the right engine torque rose to around 100%. Simultaneously, the left engine propeller rpm fell from the 100% initially achieved by both propellers to 57% whilst other engine shaft speeds (Nh, Nl) remained approximately at their originally achieved values. At this time, the First Officer called out the newly increased left engine torque as 122% and then 126%.
A few seconds later, half a minute after takeoff power had been set, the First Officer called “okay speed alive sixty” and the Captain followed this with “do we reduce or” and the First Officer responded with “we can just cut”. When the Captain said “do we abort or continue?”, the First Officer responded by calling out the engine torques - 112% (left) and 94% (right). Further torque callouts were made in response to the Captain asking “how is it now?” of 102% and 94% and then a few seconds later 116%, 94%.
After one minute from the setting of takeoff power, the Captain was evidently aware that the aircraft was not accelerating at the usual rate and was recorded saying “is it really going?” and received the response (re the engine torques) of “101, 95” which he acknowledged and almost at once again queried “is it going really, is the aircraft really moving”. Again the First Officer with the (unchanged) engine torque readings of “101, 95”. This time, the Captain responded with “it is not giving power” and the First Officer announced “okay speed has now reached about a hundred” following this with the torques “111, 95”. Approximately 1 minute and 18 seconds after takeoff power had been set, the First Officer called “a hundred now” and after a further 8 seconds “V1, Rotate” and the aircraft became airborne, according to ATC radar “near to the end of the (4,417 metre-long) runway”.
Over a period of about three seconds, the First Officer was then recorded as calling “positive rate of climb” and the Captain responding with “gear up”. A few seconds later, the Captain said “it doesn’t have power [pause] it’s on one side” with the First Officer responding with “we can also turn back”. The first of seven GPWS Mode 3 “DON’T SINK” Alerts was then annunciated and the remainder followed as the flight struggled to climb. After the third of these, the First Officer said “we can turn back” but the Captain then immediately responded with “let’s just go” and the First Officer replied “okay”. After the fifth one, the Captain said “we can turn back?” and after the sixth Nairobi TWR called to transfer the flight to Radar but after receiving no response repeated the instruction and the Capt said “tell him [pause] tell him we have no power” and in the last recorded CVR speech, the First Officer began the call but the transmission suddenly ended “along with simultaneous sounds of distress”.
According to FDR data, at no time did the airspeed exceed 102 KCAS and after an initial post takeoff climb of 120 feet above runway level, altitude then decreased by 60 feet before increasing by 50 feet. Also, immediately after it became airborne, the aircraft had begun a slow but continuous drift to the left away from the extended runway centreline. The accident site was just over 1 nm north northeast of the departure end of runway 06 coincident with an unoccupied single storey commercial building adjacent to a major road. The aircraft was found to have collided with the building at an angle and when inverted and a fierce fuel-fed fire had immediately broken out which destroyed both the building and the aircraft. The flight time from becoming airborne until the end of FDR data was 51 seconds.
The Performance of the Flight Crew
No evidence was found that physiological factors or incapacitation had affected the performance of either pilot. The occurrence of a continuous series of ‘Triple Chime’ Warnings, which signal the highest level 3 crew alert and began well before V1 was reached, was completely ignored by the crew when an immediate rejected takeoff was mandatory. This audio alert would have been accompanied by flashing red Master Warning Lights centrally positioned in front of each pilot and another flashing red light on the Central Annunciator Panel (CAP). Since the origin of these warnings was unlikely to have been either Takeoff Configuration or Landing Gear up when Flaps at the landing setting, which are the only two sources of a level 3 Alert that can be cancelled, it is of note that this repeated warning was not cancelled.
The Training of the Flight Crew
The Investigation was unable to determine whether either pilot had ever demonstrated their ability to fly the aircraft with one engine inoperative during training or otherwise. Available training records were annotated with ‘In Sim Only’ against engine failure training and “no evidence was found to show that the crew actually underwent any simulator training”.
The Airworthiness of the Aircraft
There was evidence that whatever was wrong with the left engine - evidenced from both CVR and FDR data - it was not the first time it had indicated a malfunction of some sort. An identical level 3 alert was recorded for the short positioning flight made the previous day but the First Officer on that flight denied any knowledge of it despite the fact that the crew “spent considerable time trying to troubleshoot the cause”. Other level 3 alerts had also occurred previously but there was no record of any related Aircraft Technical Log defect entry or relevant maintenance action. More generally, “there was no evidence of any maintenance having been conducted on the aircraft since its Certificate of Airworthiness issue two months previously”. Unfortunately, the condition of the engine after the accident meant that it was not recovered for examination.
The Takeoff Weight of the Aircraft
According to the information about the loading of the aircraft held at Despatch, the actual TOW (takeoff weight) was 20,167 kg including 5000 kg of cargo and was within the MTOW of 20,820 kg. However, given that the weight of cargo recovered from the accident site was 5,520 kg and that there had been unauthorised removal of some of the cargo from the accident site before it was collected together to assess its weight, it was concluded that the actual aircraft takeoff weight had exceeded the maximum permitted by an amount which could not be determined. This meant that the possible effect on the handling of the aircraft attributable to that cause could not be established. A review by the Investigation of the process by which cargo of the sort which had been carried was received and prepared for aircraft loading was inconclusive and it appeared that even Air Way Bills (AWB) were not prepared on a per flight basis. In any case, ATC confirmed that three of the operator’s cargo only flights had departed Nairobi on the day of the accident but copies of only two AWBs were available.
The Probable Cause of the accident was determined as “the decision by the crew to conduct the flight with a known mechanical problem and their failure to abort or reject the takeoff after receiving twenty seven cautions”.
The Final Report of the Investigation was completed on 26 November 2019 and subsequently published. No Safety Recommendations were made and no resulting Safety Action was recorded.