Perte de contrôle après le décollage

Extracts from final report :

Vol régulier AD Tioumen - AD Surgut (Fédération de Russie)

On 02.04.2012, at 01:35 UTC (07:35 local time), at day time, under VMC after the takeoff from the Roschino (Tyumen) airport RWY 21, the ??R72-201 VP-BYZ aircraft, operated by JSC "UTAir Aviation" (further referred to as "UTAir") crashed while performing the scheduled passenger flight UTA120 from Tyumen to Surgut. According to the load sheet the A/C TOW and center of gravity were 18 730 kg and 30.72 % MAC correspondingly and that was within the aircraft operation limits. Onboard there were 4 crew members (PIC, F/O and two flight attendants) and 39 passengers, all RF citizens.After the landing gear and the flaps retraction the aircraft started descending with a significant left bank and then collided with terrain. The ground collision first led to the structural damage of left wing followed by the fuel spillage and fire, and further to the complete destruction of aircraft with the right wing, cockpit and rear section with empennage separation.Out of the 43 persons onboard, 4 crew members and 29 passengers were killed. Others received serious injuries.

On the basis of the analysis of the found out facts and the circumstances of the flight, the results of the field investigation, including the wreckage plot of the accident site made with the use of unmanned aerial vehicle, the data of the ground and flight recorders, engineering simulation of the flight performed by the ATR Aircraft Manufacturer, the analysis of the air navigation and meteorological support of the flight, the results of the special MFC examination; the given data on the crew qualification and the organization of the flight operation within the airline; the medical data; the information of aircraft ground handling and maintenance; the results of the simulator session;, the assessment of the crew actions; the results of the forensic expertise it has been established that:

3.1.1 The ??R72-201 VP-BYZ aircraft had valid registration and airworthiness certificates. Before takeoff from Tyumen airport the aircraft was serviceable, its components and engines had sufficient life times and time limits to perform the flight.

3.1.2 The aircraft was fuelled with a sufficient quantity of fuel for the planned flight route, taking into account the chosen alternate airdromes. According to the results of the performed examinations there are no remarks as to the physicochemical quality indexes of the fuel.

3.1.3 The TOW (18730 kg) and CG (30.72% MAC) of the aircraft did not exceed the limitations, set by AFM.

3.1.4 The investigation team did not find any failures of the airframe, engines and systems functions till the collision with ground. There was no in-flight aircraft fire or destruction.

3.1.5 The meteorological support of the flight was in compliance with the regulatory documents provisions. Actual weather conditions at takeoff: wind 240 06 m/sec gusts 9 m/sec, visibility 10 km, broken clouds (5-7 oct.) cumulonimbus, cloud base 390 m, temperature -1°?, dew point -1°?, QNH 1002 hPa did not obstruct the flight dispatch.

3.1.6 The day before the accident the descent to approach for landing at the Roschino aerodrome was performed in multi-layer clouds under icing conditions, which does not exclude the residual in-flight icing accretion. During the aircraft stand (more than 7 hrs), under the atmosphere front passage, at the Roschino aerodrome there was precipitation with the near zero temperature and strong wind contributing to the ground icing accretion The right side of the aircraft (RH wing and RH horizontal stabilizer) must have been subject to more intensive ground icing as the situated windward during the period of most intensive fallout of precipitation.

3.1.7 Due to a presence of non-removed ground icing on wing, fuselage and horizontal stabilizer surfaces the aircraft aerodynamic performance at accident takeoff did not meet the aircraft type characteristics. The actual value of the lift coefficient as compared to the type values decreased by ?25%, with drag coefficient practically doubled.

3.1.8 There were timely transferred alerts of the presence of the ground icing conditions at the aerodrome. The alerts were transferred to all relevant services, including the PPCD dispatcher of UTAir-Technic, whose personnel was in charge of the aircraft maintenance and ground handling, and UTAir office. There was no mandatory procedure in the current regulations to inform a flight crew about the abovementioned alerts during a pre-flight weather briefing however the abovementioned documents were available and, if interested, flight crew members might have studied these documents.

3.1.9 Federal aviation regulations «The preparation and conducting of flights in civil aviation in the Russian Federation» (FAP-128, item 2.14) prohibit flight crews (pilots) from starting a flight, if there is frost, slush or ice on the wings, fuselage, control surfaces, tailplane, propellers, powerplants surfaces, windshields or ports of the pressure instruments - that is the requirement of the clean aircraft concept is introduced. The same requirements can be also found in the ATR 72 AFM.

3.1.10 At the time of the accident there was no basic regulatory document in force in the Russian Federation as far as the aircraft ground icing protection is concerned. The operators worked out the aircraft ground icing protection programs autonomously.

3.1.11 The provisions of UTAir documents (Flight Operations Manual, Ground Handling Management Manual, Anti-icing policy), that determine anti-icing procedures, in general met the provisions of the ICAO Manual of Ground De-Icing/Anti-Icing Operations, and also the provisions of the domestic and international guidelines. The abovementioned documents were available for UTAir-Technic who in accordance with the contract provided maintenance and ground handling to UTAir A/Cs.

3.1.12 According to the existing international practice, the aircraft de-icing/anti-icing is not considered a part of maintenance that can be carried out by certified aviation specialists only. This kind of works is related to ground handling. As de-/anti-icing treatment works are part of A/C operations and affect the airworthiness so these works must be done by trained, qualified and licensed personnel that has been trained in the certified training center. An operator shall provide the Quality Assurance.

3.1.13 The separation of the conceptions of ground handling and maintenance within the aircraft operation is not fully reflected in the regulatory documents for civil aviation in the Russian Federation. The de/anti-icing activities are not sorted out as far as the aircraft condition evaluation, decision-making on aircraft need to be de-iced/anti-iced, and its quality control and responsibility are concerned. This issue is not cleared even within major airports. The PIC who makes decision on departure and is, finally, responsible for Clean Aircraft Concept compliance does not usually have the physical possibility to inspect highly located aircraft construction elements (wing, stabilizer), moreover he cannot adequately evaluate the quality of de-icing/anti-icing treatment staying in a cockpit. From the point of view of the regulatory documents it is necessary to clearly separate the functions and responsibility of the maintenance, ground handling and flight personnel as far as the compliance with clean aircraft concept is concerned.

3.1.14 The pre-flight ground handling of the ATR 72 VP-BYZ aircraft was carried out by the UTAir-Technic personnel that is authorized for the ATR 72 aircraft ground handling only. In violation of the provisions of UTAir anti-icing policy and UTAir-Technic ATR 42/72 de-icing/anti-icing procedures the certified maintenance staff did not assist and did not monitor the quality of the ground personnel work.

3.1.15 The training of the UTAir-Technic ground handling personnel in terms of de-icing/anti-icing procedures was carried out in accordance with the Personnel training program on the conduct of ATR 42/72 aircraft after-arrival, on-stand and before-departure service (for the airports where there are no certified personnel), approved by ROSTRANSNADZOR Civil Aircraft Continued Airworthiness Department and agreed with UTAir. Only 30 minutes of lectures are dedicated to study de-icing/anti-icing issues. The content of the Program does not allow training the specialists in a proper manner, especially if they do not have aviation background.

3.1.16 The UTAir-Technic personnel training was performed by UTAir-Technic itself and by the Tyumen airport, which both are not certified training organizations and that violates the UTAir Ground Handling Management Manual. UTAir didn't provide oversight (audit) of the UTAir-Technic (as a subcontractor) activities and on personnel training.

3.1.17 Practically the implementation of the safety management system in UTAir is at the initial phase. Airline SMS approved by FATA Flight Safety Inspection Head and Airline Director General and agreed by Tyumen FATA ITO AT specialists contains general issues mainly and is not adopted for the implementation of Airline activities in particular areas. Different airline documents that determine the safety management system need to be reviewed and mutually coordinated. At the time of the accident the safety management system was inefficient and had not become the guideline for the ordinary staff.

3.1.18 The crew members had valid aviation specialist's licenses.

3.1.19 The crew members had valid medical certificates. The accident has no attitude to the crew members' state of health.

3.1.20 Under the maximum possible flight crew workload in 2011-2012 the Airline permitted violations of their work and rest balance, in particular there were 5 split flight shifts in March 2012 when only 2 split flight shifts are allowed by regulations. The PIC had 111 unused days-off, and the F/O had 123 unused days-off. Following the results of psychological team analysis the conclusion was made that with great degree of probability the crew performed the flight with background accumulated fatigue. The fatigue accumulation provokes the symptoms of increased fatigability, inattention, distraction and, as a consequence, committing errors in piloting and decision-making.

3.1.21 According to the given documents, the crew members underwent all the ATR 72 training, prescribed by regulatory documents, including the training on aircraft stall prevention including under the conditions of possible in-flight icing. PIC transition and classroom training was performed by Tyumen Personnel Training Center, NPP. FO transition training was performed in Tyumen Personnel Training Center, NPP in accordance with Sabenavita (Lithuania) Training Center program; his further classroom training was performed by Tyumen Personnel Training Center, NPP. The flight crew members' simulator training was performed in Toulouse (France) and Helsinki (Finland).

3.1.22 The manufacturer guidelines on Cold Weather Operations are issued in English only. In the Russian Federation there are no qualification requirements concerning the English language proficiency for pilots that fly aircraft with operation documentation in English only. According to expert conclusion based on tests analysis that were passed by pilots in Tyumen Personnel Training Center, NPP the PIC and F/O language proficiency did not allow them to understand to the full extent the content of the methodological technical documentation.

3.1.23 The ATR72 VP-BYZ aircraft ground handling was assigned to the aircraft mechanic who does not have aviation background. The aircraft mechanic was authorized to his duties only under the shift engineer (head) supervising. He didn't obtain the authorization to release the A/C to flight.

3.1.24 Before the servicing of ATR72 VP-BYZ the aircraft mechanic performed the servicing of another A/C - ATR72-500 VQ-BLI, including de-icing/anti-icing treatment as well as job-card drawing and signing.

3.1.25 The evaluation of the ATR72 VP-BYZ aircraft condition was performed visually by the aircraft mechanic, from the ground, the upper wing and stabilizer surfaces were not inspected. From the ground, without the use of step-ladders it is impossible to check the conditions of upper surfaces of wing and stabilizer to full extent. However based on the results of the inspection the aircraft mechanic reported to the PIC that the aircraft was clean. The PIC was not informed that the upper wing and stabilizer surfaces had not been inspected.

3.1.26 Regardless of the de-/anti-icing treatment that was being performed for the next stands aircraft the PIC performed the pre-flight inspection in a pro forma manner without paying adequate attention to the presence of snow-and-ice deposits on critical surfaces of A/C (wing and horizontal stabilizer), possibly hoping that the ground staff report on the aircraft condition would be objective.

3.1.27 In the presence of show-and-ice deposits on the ground and environment objects, in violation of FAP-128 (item 2.14) and ATR 72 AFM requirements that prohibit the flight in presence of show, ice or frost on A/C critical surfaces, as well as of UTAir Program for De-/Anti-icing A/C treatment providing that stipulates the mandatory de-/anti-icing treatment to an A/C in such conditions, the PIC and the aircraft mechanic had made a groundless but agreed decision that there is no need for the A/C to be de-iced/anti-iced.

3.1.28 As the reason of release the A/C to flight without de-/anti-icing the aircraft mechanic reported to the shift head the following: the PIC's refusal of the A/C de-/anti-icing treatment. The shift head though being aware of the aircraft conditions after a long-term stand (as he had already treated several aircraft), did not demand the execution of his earlier given instruction on the de-/anti-icing treatment of each aircraft, didn't reported to the Production and Dispatcher Service and didn't stop the flight of the accident A/C. In fact, the shift engineer kept himself aloof of the aircraft mechanic activities monitoring.

3.1.29 To provide aircraft de-/anti-icing treatment in the AP three special purpose vehicles were engaged, there was anti-icing fluid in a sufficient quantity. De-icing/anti-icing vehicle had come at the ATR 72 VP-BYZ aircraft stand on time and left no earlier than after the aircraft mechanic signal and the information about the PIC's refusal from the de-/anti-icing treatment.

3.1.30 Within one and a half hour before and after the accident except ATR 72 VP-BYZ there were nine more aircraft taking off from the Tyumen airport (?319, B737, ATR42 and ATR72), 4 of them operated by UTAir. Excluding transit flights (Sibir and Aeroflot airline), all aircraft underwent de-/anti-icing.

3.1.31 In the course of taxi and takeoff, up to onset of the emergency, the crew, all in all, performed the SOP correctly. As a shortcoming it should be noted that the elevator trim was set by the crew in position -1.65° (nose up), which was not equal to the estimated value of -0.7° (nose up) for the actual CG.

3.1.32 In the course of taxiing the PIC gave the command to the F/O to activate aircraft anti-icing system in de-icing mode, after that the crew observed the ice and snow contamination fall. The de-icing activation of aircraft anti-icing system continued for 6 min 30 sec, after that the system was switched off by the crew.

3.1.33 The crew did not consider the decision to stop taxi and return for de-/anti-icing. The crew, the PIC in particular, manifested the unawareness of the ground icing danger that, on the investigation team's opinion, is connected to shortcomings in both initial training programs (within higher school for civil aviation) and the transition course and recurrent training for ATR 72 type as far as the ground icing influence is concerned.

3.1.34 Takeoff was performed with flaps extended at 15°. The first level of the anti-icing system (the heating of sensors and windshield) was activated.

3.1.35 The aeronautical service of the flight complied with the regulatory documents in force. The actions of the ATC, the flights radio technical service and aeronautical telecommunications have no relation to the causes of the accident. The aircraft departure was performed by schedule. The runway was completely vacant; there were no time constraints for the crew for pre-flight and departure preparation. The time interval after the previous aircraft takeoff (B 737) amounted to 1min 30 sec that is in compliance of the regulatory documents requirements. The causes of the accident are not related to the wake turbulence.

3.1.36 The climb after takeoff was performed at significantly higher angles of attack that was caused by the significant decrease of the wing lift effectiveness. There is no angle of attack indicator aboard so the crew didn't have a possibility to monitor this parameter.

3.1.37 During the climb-out the crew had to trim the aircraft intensively in nose down direction almost to the structural limit. The need to trim is explained by the redistribution of the aerodynamic forces (the change of the hinge moment) along the elevator surface due to the contamination of the stabilizer upper surface. The aircraft documents (AFM, FCOM) do not contain this effect description and the crew recommended actions. At the same time this exercise can be performed on the simulator. The Airline didn't provide the Investigation Team with the data of the crew simulator training on takeoff with contaminated stabilizer.

3.1.38 Till the moment of the flaps retraction, there were no other obvious signs that could be the evidence of the "unusual" aircraft behavior, except increased angle of attack values and the redistribution of forces on the elevator.

3.1.39 At the altitude of 640 ft the autopilot was engaged.

3.1.40 During takeoff the crew performed all the actions at speeds that are recommended for the normal flight on "clean" aircraft.

3.1.41 The flaps retraction was initiated at the attitude of 640 ft at speed of 139 kt (the reference speed under normal conditions is 132 kt, under possible (in-flight) icing conditions it is 160 kt).

3.1.42 At the moment of the flaps retraction end the aircraft self-induced bank occurred due to the start of the stall. The bank appeared at the operational angles of attack, with engaged autopilot, before the activation of stall warning system. Before that moment the crew noted the buffeting.

3.1.43 Neither this time, nor later, after the activation of stall warning system (cricket and stick shaker), and even after the pusher activation the crew did not recognize the stall of the aircraft, and, consequently, did not take measures on stall recovery (to push the control column and to extend flaps); the stick pusher was overpowered by the crew.

3.1.44 The analysis of several accidents circumstances, connected to the aircraft stall, occurred in domestic and foreign civil aviation showed, that the existing methods of the simulator training on stall recognition are not effective enough. In majority of cases, in real flights, the crews that underwent simulator training failed to identify the stall and did not take corrective measures on stall recovery, and again this prove the importance of all necessary measures to prevent the A/C to run into the such flight conditions.

3.1.45 Actual actions of the crew after aircraft stall consisted in an attempt to counteract the bank (the right initially, then the left one) by the control wheel deflection. The control column was deflected to nose up.

3.1.46 The simulator session showed that without applying the recommended procedure on stall recovery (to push the control column and to extend flaps) the recovery to normal flight was impossible. If the crew had complied with the recommended procedure the altitude loss for recovery, most likely, would have been within 300-400 ft.

3.1.47 The aircraft collision with the ground occurred at the pitch angle ~11° nose down, left bank of about 55° and vertical speed of descent more than 20 m/sec, which predetermined severe destruction of the aircraft and the death of most people aboard. After the collision with ground the post-impact fire rose on certain elements.

3.1.48 Search and rescue works, all in all, were performed effectively, that allowed to evacuate surviving people on time and ensure the accident site guard.

3.1.49 There were no aircraft construction features revealed that might have had negative influence on the crew and passengers survival.

3.1.50 There were no information of COSPAS-SARSAT emergency beacon activation reported from International Coordination Computer Center (Moscow, RF) to Ural Search and Rescue Coordination center (Uralsky APSC). The emergency beacon was not registered in the International Coordination Computer Center (Moscow, RF).

The immediate cause of the ??R 72-201 VP-BYZ aircraft accident was the PIC's decision to takeoff without de/anti-icing treatment despite the fact that snow and ice deposits were present on aircraft surface and were discovered by the crew members during taxi which resulted in degradation of aircraft aerodynamic performance and stall during climbing after takeoff as well as inability of the crew to recognize stall and, consequently, failure to undertake recovery procedure.

The aircraft stall occurred at the operational angles of attack right after flaps retraction with engaged autopilot before stall warning system activation and was caused by the loss of the wing lift effectiveness due to takeoff with non-removed ground icing.

The system cause of the accident were shortcomings in ground handling activities and staff training in UTAir-Technik that became possible because of absence of due monitoring by the Technical and Operation Supervising Directorates of UTAir airline for compliance with airline requirements regarding ground handling and aircraft ground icing protection which resulted in erroneous evaluation of aircraft conditions by the PIC and aircraft mechanic (the shift head kept himself aloof from monitoring mechanic's activities) after the aircraft has been on ground in icing conditions for a long time and in release the aircraft to fly without de/anti-icing treatment.

non-execution of measures on de-icing was the consequence of the erroneous coordinated decision of the PIC and the aircraft mechanic who was in charge of the aircraft pre-flight servicing after long aircraft stand under conditions, contributing to the ground icing accumulation of the aircraft. There was no control of the aircraft mechanic work on behalf of the shift engineer.

The departure of the aircraft with non-removed ground icing became possible due to system shortcomings in the aircraft ground icing protection organization in the Russian Federation, that at the time of the accident was revealed in the absence of the valid basic document, that would determine the regulatory requirements as to aircraft ground de/anti-icing, including the requirements for organizations licensing and training of the personnel that is responsible for the works in question.

The contributing factors were:

· The shortcomings in the UTAir safety management system which contains, all in all, general issues only and is not adopted for the implementation of Airline activities in particular areas, which did not allow to reveal and correct existing safety risks in a timely manner.

· The shortcomings in the UTAir-Technic quality management system, resulted in neglecting of certain requirements of the UTAir ground handling management manual regarding staff training and monitoring for aircraft de/anti-icing treatment which led to the situation when not sufficiently-qualified staff performed the evaluation of the aircraft surface conditions and made the decision on need for the aircraft to be de-iced/anti-iced.

· The absence at the time of the accident of regulations that establish state requirements for ground handling (de/anti-icing treatment in particular) including staff training and organization licensing.

· The shortcomings in crew members initial and recurrent training as far as the danger of ground icing, its influence on the aircraft aerodynamic performance together with aircraft anti-icing system operation features and design are concerned that did not allow the crew to make the only appropriate decision to return for de-icing/anti-icing treatment after the observation of the snow and ice contamination on the wing after anti-icing system activation in de-icing mode while taxiing for takeoff.

· The methodological imperfection of the crew computer based and simulator training programs concerning the prevention of aircraft stall, identification of approach to stall and taking timely actions for recovery.

· The increasing need for number of flight crews to perform highly growing flights schedule which, with ineffective SMS, resulted in flight instructor work deficiencies during PIC training and absence of PIC skills to take correct decisions and to strictly comply with the regulations in force.

· The possible fatigue of the crew members due to the violation of the work and rest balance while performing split flight shifts together with a large number of unused days-off.