D228, vicinity Kathmandu Nepal, 2012
From SKYbrary Wiki
|On 28 September 2012, control of a Sita Air Dornier 228 being flown by an experienced pilot was lost at approximately 100 feet aal after take off from Kathmandu in benign daylight weather conditions and the aircraft stalled without obvious attempt at recovery before impacting the ground where a fire broke out. All occupants were killed and the aircraft was destroyed. The comprehensive investigation found that insufficient engine thrust was being delivered to sustain flight but, having eliminated engine bird ingestion and aircraft loading issues, was unable to establish any environmental, airworthiness or loading issue which might have caused this.|
|Actual or Potential
|Human Factors, Loss of Control|
|Aircraft||FAIRCHILD DORNIER 228|
|Type of Flight||Public Transport (Passenger)|
|Origin||Kathmandu/Tribhuvan International Airport|
|Intended Destination||Lukla/Tenzing-Hillary Airport|
|Take off Commenced||Yes|
|ICL / ENR|
|Location - Airport|
|Airport vicinity||Kathmandu/Tribhuvan International Airport|
|Tag(s)||Root Cause Not Determined|
Inappropriate crew response (technical fault),
Plan Continuation Bias
|Tag(s)||Loss of Engine Power,|
Bird or Animal Strike,
|Tag(s)||Airport Emergency Medical Response,|
|Damage or injury||Yes|
|Aircraft damage||Hull loss|
|Fatalities||Most or all occupants (19)|
|Causal Factor Group(s)|
On 28 September 2012, a Dornier 228-200 being operated by Sita Air on a domestic scheduled passenger flight from Kathmandu to Lukla failed to sustain its initial climb after take off in day Visual Meteorological Conditions (VMC) and having reached a maximum height of about 100 feet aal, it entered a stall from which recovery was not achieved before ground impact and subsequent fire which led to the death of all 19 occupants and the destruction of the aircraft.
An Investigation was carried out by an Aircraft Accident Investigation Commission appointed by the Nepalese Government. The Flight Data Recorder (FDR) and Cockpit Voice Recorder (CVR) were recovered and downloaded but some FDR information was missing because it relied on sensors external to the recorder which had been disabled by the effects of the impact and fire. The CVR recording was used undertake Spectrum analysis of the CAM channel to analyse the sound of each engine as an “audio signature” and to identify the incidence of other noises, including one which corresponded with a ‘flash’ captured on CCTV appearing to come from one of the engines shortly before the aircraft became airborne.
With the type-experienced aircraft commander acting as PF and accompanied by an inexperienced but correctly qualified co pilot, the aircraft made an intersection take off from the 3050 metre long runway 20 which provided an available runway length of 2112 metres. The aircraft accelerated normally until approximately 70 knots - well below the applicable V1/Rotation Speed (Vr) speed of 83 knots. At this speed, at about the same time as the aircraft hit a Black Kite (which would have had an estimated live weight of around 700 grams), a wrong call of “V1 Rotate” by the co pilot was followed by a brief but unsuccessful attempt to rotate the aircraft. The entire carcass of the bird was subsequently recovered from the runway and the Investigation found no evidence of any bird remains in either engine. Continued acceleration on the ground above 70 knots was less than normal but just before the aircraft became airborne, a small ‘bang’ or ‘thud’ can be heard on the CVR CAM channel which corresponds to the signatures associated with the engine rotation rate of one of the engines dropping by approximately 4% before recovering. At approximately 86 knots, which was above the applicable V1, Vr and V2, the aircraft had become airborne and began to climb away.
The Investigation found that CVR audio signatures associated with the right engine had remained constant at 100% while the aircraft remained airborne but those for the left engine dropped to 95% of their nominal frequencies just after the landing gear was raised, and dropped further to 91% shortly afterwards. The aircraft reached a speed of 89 knots in the two seconds after becoming airborne but speed then dropped to 77 knots as it climbed to 100 feet aal. At this point, there was no longer sufficient thrust from the engines to overcome the drag of the aircraft and further speed loss continued as the aircraft drifted to the left of the runway centreline until successive stall warnings were activated and it then stalled from a left turn and, having departed controlled flight, impacted the ground in an extreme nose-low attitude. It was established that both engines had been operating and delivering low or flight idle power at the point of impact and that the propellers had not been feathered.
The Investigation established that when the engines of the accident aircraft had been installed, the mandatory Aircraft Maintenance Manual check of Flight Idle fuel flow with the Speed Lever set to low during an engine ground run had not met the minimum permissible figure of 96.5 ±0.5%. Instead, the recorded figure had been 90%, within the Operator's own engine ground run performance check requirement for the flight idle fuel flow check to result in an engine rpm of “approximately 88%”. It was also found that there was no record of any airborne flight idle fuel flow check having been carried out by the operator in accordance with the procedure detailed in the Pilots Operating Handbook (POH). It was noted that “this lower than required rpm would result in a lower than required fuel flow rate which would result in higher drag from an engine at flight idle” and that “it could also result in the in-flight flight idle rpm dropping below 100% rpm with the Speed Lever remaining in high”. However, it was not possible to determine the potential effect of this on the trajectory of the accident aircraft.
It was found that pilot training in aircraft handling, including loss of engine power and asymmetric flight, was carried out on the aircraft rather than in full flight simulator, so that control of an aircraft near V1 on the ground or at low altitude after take off could not be practised.
The actual take off weight of the aircraft was calculated, using standard passenger weights, to have been 5914 kg whereas the load and trim sheet completed and accepted prior to the flight showed 5834 kg, just within the RTOW for the flight of 5,836 kg - less than the 6200 kg MTOW for the aircraft because of the effect of density altitude on aircraft performance due to Kathmandu airport being at an altitude of 4390 feet amsl. However, after comparing the actual climb performance of the aircraft with that achieved, the Investigation was satisfied that “overloading alone was not the cause of the lack of performance in the climb phase”. Also, whilst it was found that the trim sheet for the flight had recorded incorrect positions for the aircraft centre of gravity, it was established that this had not resulted in an out-of-trim condition.
It was noted that the response of the airport Rescue and Fire Fighting Services to the crash site was delayed because “Army personnel were blocking the route to the exit and personnel manning the exit required permission to unlock the gates to let the fire vehicles pass”. It was also noted that “once on public roads, the vehicles were delayed further by the weight of traffic and a crowd of people near the impact site”.
The Investigation was unable to determine the cause of the low power which led to the loss of control but did identify the following Causal Factors:
- During (the) level flight phase of the aircraft, the drag on the aircraft was greater than the power available and the aircraft decelerated. That resulted in excessive drag in such critical phase of ascent lowering the required thrust. The investigation was unable to determine the reason for the reduced thrust.
- The flight crew did not maintain the airspeed above the stall speed and there was insufficient height available to recover when the aircraft departed controlled flight.
In addition the following Contributory Factors were also identified:
- The flight crew did not maintain V2 during the climb and so (that) the power required to maintain level flight was greater than it would otherwise have been.
- The flight crew did not maintain the runway centreline which removed the option of landing the aircraft on the remaining runway
A total of fifteen Safety Recommendations were issued as a result of the findings of the Investigation all aimed at supporting a general improvement in the safety regulation and management of aviation in Nepal.
Three Interim Recommendations were issued on 29 November 2012 as follows:
- that the Operators and the Civil Aviation Authority of Nepal must ensure that their respective weighing machines at the check in counter of all the airports are calibrated regularly by the appropriate authority of the Government of Nepal.
- that the Operators and the Civil Aviation Authority of Nepal (CAAN) must ensure that all the hand bags and checked baggage, especially when carried to and from STOL airports, are weighed and tagged with their weight mentioned in the respective tags before they are boarded, and load and trim sheets of all the departing aircrafts are prepared by qualified CAAN-authorised airline personnel.
- that the Flight Safety Standards Department at the Civil Aviation Authority of Nepal, ensures that all Dornier 228 aircraft flying in Nepal are checked to confirm that the flight idle fuel flow of their engines has been set in accordance with the AMM and POH.
A further twelve Recommendations were made at the conclusion of the Investigation as follows:
- that the Civil Aviation Authority of Nepal reviews its training requirements in relation to engine malfunctions at or near V1 to ensure that they are adequate for commercial pilots flying aircraft for which training is currently not carried out in a flight simulator.
- that Sita Airlines Pvt. Ltd reviews the policy that prevents pilots from landing on the remaining runway following an engine malfunction just after V1.
- that the Civil Aviation Authority of Nepal reviews the training requirements for airline personnel responsible for the loading of aircraft to ensure there is confidence that aircraft (is) dispatched properly loaded.
- that the Civil Aviation Authority of Nepal reviews the suitability of the average weights used for passengers on flights within Nepal and modifies them if necessary.
- that the Civil Aviation Authority of Nepal carries out periodic checks of aircraft loadsheets to ensure they have been completed correctly and accurately.
- that Sita Air Pvt. Ltd ensures that passengers on its flights are given safety briefings before departure in accordance with FOR-Aeroplanes and its Operations Manual.
- that the Civil Aviation Authority of Nepal monitors airlines within Nepal to ensure that passengers are given safety briefings before departure.
- that Kathmandu Airport reviews its bird control program, the bird strike reporting system and the bird activity monitoring system to determine whether their effectiveness at reducing bird activity can be improved.
- that Kathmandu Airport manages and controls more effectively and efficiently the habitats attractive to birds on and near to the airport.
- that Kathmandu Metropolis, Lalitpur Sub-Metropolis, Bhaktapur Sub-Metropolis and Madyapur Thimi Municipality adopt more effective measures for waste management, kanji house [Editors Note: stray animal compound], butcheries and river control in order to reduce bird activity near to the Airport.
- that the Civil Aviation Authority of Nepal should ensure that controllers in the Airport Fire Watch Tower are able to see all areas appropriate to their role in and around the airport.
- that the Civil Aviation Authority of Nepal should ensure that Airport Fire Service vehicles at the airport have clear and unhindered access to the exit gates and free movement through the gates in case of emergency.
The Final Report of the Investigation was made available in August 2013.