Aeroflot SU95 at Moscow on May 5th 2019, aircraft bursts into flames during rollout and burns down

Last Update: March 30, 2025 / 13:51:31 GMT/Zulu time

Bookmark this article

Photo of Aeroflot RA-89098, SUKHOI Superjet 100-95 — Aeroflot RA-89098, SUKHOI Superjet 100-95 (Photo credit: Anna Zvereva / Flickr / License: CC by-sa)

Incident Facts

Date of incident
May 5, 2019

Classification
Accident

Airline
Aeroflot

Aircraft Registration
RA-89098

Aircraft Type
SUKHOI Superjet 100-95

ICAO Type Designator
SU95

On Mar 29th 2025 the MAK released their final report concluding the probable causes of the accident were:

The air accident to the RRJ-95B RA-89098 aircraft was caused by the uncoordinated control inputs by the PIC at the flare, landing and through the several repeated bounces of the aircraft off the RWY (the porpoising), having manifested in the several disproportionate alternating sidestick inputs in pitch with keeping the sidestick retained against each stop. The indicated control inputs had resulted in three hard touchdowns of the aircraft, as a consequence at the second and third touchdowns the absorbed energy significantly exceeded the maximum values, for which the structural integrity had been evaluated at the aircraft type certification, which led to the destruction of the airframe structural elements, the fuel tanks with the fuel spillage and the fire onset.

The contributing factors to the accident were:

- the ineffectiveness of the RRJ-95 flight personnel approved training programs as for the actions into the major failure condition//abnormal situation at the FBWCS reversion to DIRECT MODE and, consequently, the insufficient knowledge and skills at the flight crewmembers to operate the airplane in this mode. The training programs met the minimum requirements, determined by FAR, but did not account for the specific nature of a particular emergency;

- the ineffectiveness of the airline SMS in terms of the monitoring of the piloting sustainable skills development at the pilots, which prevented the identification and elimination of the PIC’s common systematic errors at the sidestick pitch control at the stage of landing, including these, associated with its forward inputs beyond neutral (to nose down) into the flare;

- the failure to identify the biases (hazards) in the airline flight crews’ piloting technique as far the previous events of the FBWCS reversion to DIRECT MODE are concerned and thus the failure to implement preventive measures;

- the aircraft operational documentation unclear wording in terms of the piloting peculiarities at flare and the correction of the deviations at the landing (counteracting the consecutive aircraft separations off the RWY);

- the failure of the crew to comply to the FAR and OM requirements at the flight preparation and performance at the actual and forecast thunderstorm activity, as well as at the availability to observe these zones on the weather radar display, which had resulted in the aircraft encounter the atmospheric electricity, the EIUs reboot and the FBWCS reversion to DIRECT MODE. As per the certification results the FBWCS reversion to DIRECT MODE had been assessed as «the major failure condition», the in-flight onset of «the major failure condition» at the lightning or static electricity exposure does not contradict the applicable certification requirements;

- the dramatic increase of the psychoemotional stress at the PIC because of the aircraft exposure to atmospheric electricity and the failure within a long time to ensure the acceptable piloting precision at the FBWCS in DIRECT MODE, which led to the psychological dominant mindset formation to perform immediate landing together with the lack of readiness to initiate goaround (not go-around minded);

- psychological personality traits of the flight crewmembers that determine their behavior in the stress environment, as well as the PIC’s insufficient training in human factor/performance and threat and error management approach, which prevented the objective assessment of his psychoemotional condition and the ability to control the airplane, to choose the optimal strategy to proceed the flight, as well as to establish the required interaction and CRM;  the failure of the PIC to ensure the aircraft pitch trim under the manual control, including at the glideslope descent;

- the incorrect assessment of the situation by the crew at the Predictive Windshear warning (GO AROUND WINDSHEAR AHEAD) trigger at the flight on glideslope and, consequently, the non-initiation of a go-around maneuver, that resulted in the aircraft encounter the wind microburst at the early flare and affected the aircraft flight path. The documentation by the aircraft designer and the airline allows the crew to ignore the subject warning activation, if it made sure there is «no windshear threat», still the operational documentation and the OM do not integrate the respective clear criteria of «no threat»;

- the purposeful ducking under the glideslope by the PIC at the final approach (after passing DH);

- the difference between the airline OM provisions asfor the crew actions at the glideslope warning activation (the excessive deviation off the glideslope equisignal zone) and the similar provisions in the aircraft designer documentation. Subject to the provisions of the aircraft designer documentation the crew should have performed go-around;

- the unjustified extension by the airline of the approach «stabilized condition» criteria as for the acceptable deviations range off the target speed, which at the actual IAS of 15 kt higher against the target one and the FBWCS in DIRECT MODE resulted in the unexpected for the PIC increased aircraft response to the sidestick input in pitch;

- the failure by the crew to carry out the SOP on the manual speedbrakes deployment at the aircraft touchdown. The operational documentation unclear wording and the monitoring algorithms of the landing configuration, used at the aircraft that require to arm the speedbrakes for the automatic deployment, including at FBWCS in DIRECT MODE, in which the automatic deployment is disabled, degrade the crew’s situational awareness as for this aspect.

- the TR actuation after the first bounce off the RWY, which had made the subsequent go-around impossible.

As per the results of the forensic medical examination the death of 40 out of 41 fatally injured people had been caused by the exposure to open flame, accompanied with the burns of the upper respiratory tract through the inhalation of hot air.

The fire erupted after the aircraft third touchdown due to the disintegration of the wing fuel tanks and the fuel spillage. The fuel spillage occurred as due to the destruction at the landing gear retraction/extension actuating cylinders attachment points, as well as due to the destruction of the other wing parts. The landing gear structure had been damaged at the second touchdown that is at the third touchdown functioned beyond the expected operational conditions and had not been able to bear the applied landing loads as designed.

The operation (destruction) of the landing gear fuse pins («weak links») at the second touchdown had been consistent with the design integrated logic. With that the loads, actually accomplished, had been less of those in use to demonstrate compliance to AR-25 item 25.721 at the aircraft type certification, which prevented the MLG legs to completely separate off the airplane structure (it is only the Attachment A fuse pins that had been destructed). No correlation between the certification requirements for the structure, including MLG legs structure, and the conditions for demonstrating their safe separation off results in actual significant risks of the fuel tanks disintegration and the fuel spillage even in case of compliance demonstration to every single of these requirements.

At its very onset the fire by its nature had been the deflagration flash, which had been accompanied with an intense smoke release with the onset of a steady burning in two seconds. By the point of the evacuation initiation the fire had been propagated inside the cabin through a row of cabin windows at the rear fuselage along the right and left sides, with that the airworthiness standards do not set up the requirements for the cabin windows as to the external fire protection.

That situation had been beyond the expected operational conditions as there had been no time margin (90 sec), at which the crew and passengers’ emergency evacuation is demonstrated at the type certification.

Most probably the following factors had contributed to the increase in the severity of the consequences:

- the running engines of the aircraft, having been not timely shut down by the crew;

- large amount of fuel, spilling out of both wing panels, which penetrated the area of the exhaust-mixing nozzles, exposed directly to their jet streams;

- the inability to evacuate through both of the rear emergency exits;

- the manifestation of the flashover effect at the rear passenger cabin;

- the crush and panic among the passengers;

- the efforts by a number of passengers to pick up their carry-on luggage at the evacuation;

- the CFA’s error in operating the PACIS, and consequently the decline in the passengers’ situational awareness as for the evacuation procedure.

The simulation of the fire development by the Russia EMERCOM Saint Petersburg State Firefighting Service University, FSEI HE had revealed that the flight attendant’s erroneous actions to having opened the left rear door in that actual environment had not led to the increase of the fire destructive factors magnitude and had not affected the accident outcome.