Flydubai FZ981

A Boeing 737-800 attempts to land in windy weather in the small hours of the morning at Rostov-on-Don, Russia on a runway approach notorious for its windshear .

The crew fails to stabilise the aircraft on its first approach either because of windshear, or because it fails to make visual contact with the runway lights in time for a safe landing, and decides to climb away and circle, waiting for an improvement in the weather.

On its second attempt to approach the same runway – 22 – using a category 1 instrument landing system for guidance, it crashes short of the runway. There was no emergency call.

But this is no ordinary crash of the type that would have occurred if the crew – now under pressure to land because fuel is getting low – had made the decision to continue the descent through decision height, despite not being able to see the runway. If that had been true large sections of the aircraft would have remained intact.

This aircraft hit the ground about 300m short of the runway 22 threshold with such force it was shattered into tiny pieces which were scattered across the airfield. How could that happen?

Information from flight tracking service FlightRadar 24 suggests that the crew also abandoned this second approach, climbing away, but then disappearing.

On 17 November 2013 a Tatarstan Airlines Boeing 737-500, en route from Moscow to Kazan, abandoned a poorly executed night approach at its destination airport, applying full power for a go-around. The nose pitched up to 25deg and the speed rapidly dropped because of the steep climb. The crew, becoming disorientated, pushed the nose down hard, putting the aircraft into a dive at an angle of 75deg just before impact. The aircraft was shattered.

On 12 May 2010 an Afriquiyah Airways Airbus A330-200 carried out a go-around from the approach to Tripoli airport’s runway 09 at dawn, the crew lost control because of disorientation and the aircraft crashed. There was one survivor among the 104 on board.

There have been  many documented cases of crews nearly losing control when carrying out an all-engines-operating go-around.

This does not pretend to be the definitive answer to what happened to Flydubai flight FZ981 on 19 March, but it does pose the question as to what kind of event could cause the wreckage to be so badly fragmented.


8 thoughts on “Flydubai FZ981

  1. I would concur with this contemplation David, but I do hope that this scenario will prove not to be the actual cause as it would be a serious blow to safety management effectiveness of the industry as a whole.
    The fact that the aircraft had around a third of its pax seats occupied and was probably light on fuel makes go around power quite overwhelming.
    I have, in the past, read comments in the readers section in Flight International Magazine from experienced airline pilots suggesting full power go around training instead of (or in addition to) v1 cuts and n-1 go arounds, for that very same reason; risk of disorientation and/or LOC.
    Recently I had a casual conversation about this subject with an experienced pilot and TRI with plenty of airline duty under his belt and he said the same, but stated that not all programs include full power go around training. I don’t know if the sensation of disorientalion can be replicated in a simulator or whether there are other reasons not to address this in regular simulator training. Or maybe it is.. Look forward to other opinions.
    All everyone can do at this point is patiently wait for the investigation results and try to learn a lesson from this horrible accident.


  2. David, I am currently looking through some 100+ accident reports including the 2 you mention, in relation to the crew procedures aspects. A major issue seems to be the lack of preparedness of the pilot flying the go-around when it’s initiated – “plan continuation bias”.

    One way to resolve this is to split the pilot duties so that the pilot flying the approach is flying with the full intention of making a go-around; the other pilot makes the visual assessment and only takes control when satisfied that a safe landing can be made. This is of course the “monitored approach” procedure.

    Please get in touch with me if you’d like a copy of the draft paper I’ve got addressing this. Also of course the subject of a lot of material on my website as well as links to all the relevant NASA/NTSB etc research that supports this.
    Steve Last


  3. Consider a microburst encountered on short-short finals coupled with tired crew, pressure to land (being more than 2 hrs late), fuel running low for possible diversion to another airport, bad weather, poor visibility.


  4. Clearly with 4 hours fuel upon arriving there was no pressure to land. The weather wasn’t that bad with 24G36 knots down the runway, is not your everyday weather but not exactly unsafe either . WS warnings are so frequent at ROV that it is a normal day, but generally refers to rough air on approach. The weather for the second approach was technically good enough. From flightradar24, the go-around looks to be of the expected profile, but then a sudden downward trajectory. This low resolution information does not look like particularly bad decision making nor mishandling of the go-around. Regardless, they have to recorders, I preliminary announcement can be expected in a few days or weeks I should think.


    • In fact, the more I look at that go-around profile, the more it looks like a flight control failure or problem. The go-around profile is appears to be properly controlled with the initial climb and acceleration phases clearly apparent. Then the sudden dive without any change in airspeed. With the quality of that data available, there is no evidence or windshear or a stall. The request to climb to FL80 after the missed approach was clearly to remain in VMC and out of icing conditions. It is actually quite difficult to fault their decision making on the basis of the info we currently have.


  5. Looking at history (which is no proof, but an indicator of possibilities), somatogravic illusion kicks in during the go-around acceleration/climb phase, and the crew themselves do the pushing forward on the controls. That’s what happened in the Tatarstan case and others. Caveat: it doesn’t mean it definitely happened here, but it’s a possibility.


    • David,

      That would seem a very extreme reaction, especially for a second missed approach and especially considering the extreme trim and control forces required with even moderate thrust set. This is with the assumption that the aircraft remained more or less wings level throughout. The videos and trajectory seem to suggest this, but the imagery is very poor.

      In my personal experience, my worst go arounds have been when having to unexpectedly go missed. They both expected to go around and had already performed at least an adequate manoeuvre that night.

      However,as far as possibilities go, however difficult to imagine, it remains a possibility.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s