ET302, Pilots on a wing and a prayer: mindFly


ET 302 took off from Addis Ababa, the capital of Ethiopia on 10th March 2019 and 06 minutes into the flight met with an accident. The aircraft was the Boeing’s latest addition to the B-737 series, the B-737 8Max.

The infamous MCAS(software) was to have played a role in the accident and nations all over the world, fearing the safety of the traveling passengers, banned the aircraft flying in their airspace. It soon became a world wide ban, thus grounding the MAX fleet.

I will not critique on the crew action or inaction because their behaviour will need to be analysed in great depth after more details are released.

Download the Preliminary investigation report ET302 10 Mar 2019

Boeing 737 has had rudder design issues in the 1990’s when there were a series of accidents involving United, US Air and Silk Air. Read my blog on A question of Safety or Ethics:Boeing 737 mindFly . Boeing was quick to respond to it and wanted the editors to take the blog down but the evidence was over bearing and  the blog survived.


Two accidents of brand new aircrafts in a span of 4 months, roughly attributable to the same cause is a very serious issue as far as design, production and certification is concerned. The same is being investigated by the authorities in USA at different levels.

The flight path of the ill fated aircraft was mostly straight out after departure from Runway 07R. Addis Ababa airport is at an elevation of 7625feet AMSL positioned on a table top and with steep drop on either sides. There are other high elevation terrain around the airport.

Flight Path of ET302: Source Google Earth


In order to understand the events and to be able to empathise with the crew, its important to look at the accident holistically.

1.Loading of the aircraft:

ET302 is suspected to have been very close to the maximum weight that the aircraft can takeoff with.

Maximum takeoff weight limited by the environmental conditions and the thin air at the high altitude on that day was 72,400kgs.


Total weights adds to 11,910kgs as against 11,309kgs given in the load sheet presented to the flight crew. If this is true then the takeoff weight would have been 72,603kgs as against 72,400 the maximum takeoff weight and 72,000kgs for which the takeoff speeds and thrust calculated.

2. Takeoff roll:

The duration of the takeoff roll from increase in thrust to airborne as indicated by the air/ground sensors was 54seconds. This duration could be of interest to the investigators who are trying to correlate the weight/speed/thrust. In the absence of data available on the takeoff speeds it cannot be established conclusively at this stage.

3. Getting airborne:

This is the stage when things begin to go wrong. From the human factors point an element of surprise could have set in, though that can only be confirmed when the CVR transcript is available.

As the aircraft nose pitches up passing 10 degrees to the target pitch of 15 deg, 3 things happen.

  1. The Captain’s air speed indication drops by about 20-25 kts.
  2. The Stick shaker activates indicating a near stall situation.
  3. The Captain’s angle of attack sensor which was normal till 10 degrees suddenly jumps to ~35degrees and then ~74degrees nose up. The crew doesn’t have an indication of this reading in the cockpit. It is the source for activation of the stick shaker.

What happened here and why is critical for the investigation.

The Captain immediately pitches the nose down as if he was aiming to recover the lost speed and probably silence the stick shaker. The co-relation of the parameters can be seen in the image below.

Nose pitch down reaction

The crew were taken by surprise or that they knew that the aircraft was really heavy and immediately suspected the loading. Surprise is the element where the planned course of action does not materialize. Read my blog on Startle and Surprise. Till now the crew probably assumed that all aircraft systems  and protection features were operating normally.

4. After takeoff:

After getting airborne, the Captain’s Airspeed and Altimeter indications diverge from the Co-Pilot’s side, under-reading by 20-25 kts speed and ~800 ft altitude. The pilots have no indication in the cockpit that the angle of attack (AOA) sensor has probably malfunctioned or damaged.

The Question that arises here is, while the entire focus of media has been on the flight control software MCAS which automatically moves the nose down through the STAB trim, why isn’t anyone questioning the fact that in both accidents, Lion Air and Ethiopian, the problem begins with the Captain’s Angle of Attack sensor?

Climb profile

The aircraft pitch continues to be maintained less than 10 degrees till flap retraction and there after even lower and occasionally touching 0 degrees. The aircraft normally flies level with a slightly nose up pitch around 2-3 degrees depending on configuration and speed.

The aircraft climb profile is very shallow and there is a rapid increase in speed.

Autopilot is engaged briefly before it gets disconnected. The Captain asks for the speed to be selected at 238 kts showing his intention to control the rapidly increasing speed. The auto throttles coupled with the autopilot should have engaged with the autopilot but the thrust setting shows a constant 94% N1 engine thrust setting throughout the flight.

5. Automatic/Manual Pitch trim:

Automatic Pitch Trim

With the AOA sensor on the Captain’s side giving a very high unusual attitude, after flap retraction and the disengagement of the autopilot, the conditions for automatic activation of the MCAS were met. This can be seen as M1,2,3 above. The nose down trim action of the MCAS lasts for ~9 seconds. There are two bursts of manual pitch trim action by the crew before the crew realise that they need to implement the emergency procedure to deactivate the stab trim by using the STAB TRIM cutout switches.


After the STAB TRIM is disabled using the switches and implementing the FAA/Boeing procedure, the aircraft continued to accelerate with constant engine thrust towards the maximum permissible speed called the VMO.

For a good ~2min 30 sec, there is no crew action on using the manual trim to pitch the aircraft up and regain altitude or trade altitude by washing off the excess speed.

What was preventing the aircraft from climbing is the question here?

6. Reactivation of the automatic trim:

After ~2min 40 sec of the stabilizer cut out action being performed, the data reveals the activation of automatic trim for ~5sec. The reason for reactivation of the STAB TRIM is not known.The stab trim position is nearing 2 units which is close to full nose down. The stabaliser is a big control surface and much bigger than the elevator, therefore any action on the stabiliser with have a much larger pitching effect that action/counteraction through the movement of the elevator.

7. Effect of flight controls at excessive speeds:

With the speed exceeding the maximum permissible speed and clacker noise indicating the same, the aerodynamics may change. The Boeing 737 Flight Crew Training Manual is quoted below:

“Excessive airloads on the stabilizer may require effort by both pilots to correct the mis-trim. In extreme cases it may be necessary to aerodynamically relieve the airloads to allow manual trimming. Accelerate or decelerate towards the in-trim speed while attempting to trim manually”

This could be the reason why the crew were not able to manually trim the nose UP in the final moments before the crash. A full nose down trim is approximately 250 rotations of the trim wheel, this gives an indication of the effort required to pitch the nose up manually against the air loads created on the flight controls at high speed.

7. Final Moments:

The pitch trim position of the stabalizer is close to full nose down. In a moment arm, the physical movement of the stabaliser is upwards, creating lift. This action lifts the tail thereby pushed the nose down. This can be seen in the picture below, arrows pointing nose DOWN position as the mark on the top of the neutral position and nose UP as the mark below the neutral.

The elevation which is much smaller in size as compared to the stabaliser, needs to be moved up by pulling the control column back for an aircraft nose up pitch and down for nose down.


In the final moments, the last activation of the STAB trim NOSE DN pitches the nose down to -20 degrees and momentarily the control column is also moved forward. The aircraft pitches down by almost 20 degrees. Thereafter, probably both pilots move the control column back to prevent the nose down pitch. 

This action of pulling the elevator up to try and pitch the nose up creates an aerodynamic situation where both the Stabaliser and Elevator are pointing up creating a shallow “V”. In this unique position the airflow is disturbed over the tail plane and aggravated due to very high speed, the tail plane stalls. The actual position of the STAB would have been slightly more UP than shown in the picture below.

A probable position

A tail plane stall at high speed can lead to a pitch down. The aircraft pitched down to -40 degrees and the speed increased to over 500 kts. Much higher than the design speeds.

Past accidents due to horizontal stabilizer malfunction.

The short segment lasting about 6 minutes created some surprises, raised a lot of questions about the aircraft design and components. There were segments where one would feel that the crew had the aircraft under control but at the end the brut power of the engines roaring at takeoff thrust continued to accelerate the aircraft and created unique situations from where the crew unfortunately could not recover.

The first person to arrive at the scene of the accident is usually the crew. In their absence, the truth would have to be derived by piecing all possible evidence systematically. I hope the investigation is fair and unbiased so that the aviation community benefits from the learnings.