The investigation report is now out and reveals startling similarities to three other infamous crashes: British Midland Flight 92, Airfrance Flight 447 , and Pinnacle Airlines Flight 3701.
This is what happened:
On 27 January 2020, the aircraft was in a slow climb from 42,000ft to 43,000ft on an airborne communication support mission over Kabul when the left engine suffered a catastrophic failure. In a loud bang and severe onset vibrations, a fan blade separated and was ingested into the engine core.
The FADEC, detecting the failure, automatically shut the left engine down. This fully automated action, controlled by the computer and designed to prevent further damage and assist the crew, was in fact partially to blame for their final demise.
Pilots trained all the time to recognised engine failures using a combination of engine parameters, EICAS messages, and aircraft yaw sensations. What would a pilot expect to see after catastrophic engine damage? A yaw towards the failed engine. A lost of N1 speed indication. A “L ENG FLAMEOUT” message.
Yet, this crew had initially none of the above.
After the initial startle effect of the loud bang (captured on the CVR) and aircraft severe vibrations, the crew had only two messages “L FADEC FAIL” and “ENG SYNC FAIL” which could easily be interpreted as a computer failure. It took up to 40 seconds for the “L ENG FLAMEOUT” message to appear as the N2 slowly spooled down below 35%. A lot can happen in 40 seconds.
Meanwhile, the crew was working to identify the troubled engine. Without a clear EICAS message, they were left with instrument parameters and yaw sensations to identify the failed engine.
Anyone who has tried to read instrument gauges while experiencing severe turbulence understands the challenge. On that faithful day, due to a damaged sensor, the N1 indications were incoherent (from 7.6% to 255.9%) which was assumed by the crew to be related to the FADEC failure.
What about the vibration indications? When the whole plane started shaking, engine high vibrations were displayed on both engines, making it difficult to distinguish the culprit. In fact, the vibrations were so strong that it triggered the crash switch of the CVR (designed to preserve data in case of a crash), which subsequently stopped recording.
Maybe the initial yaw movement would have helped? Unlikely. If initially the plane would have yawed left, the startle response to the emergence combined with a disconnected autopilot and autothrottle was to immediately reduce thrust, making the aircraft yaw back to the right.
So the crew took matters into their own hands and applied asymmetric thrust to identify which engine was the source of the vibrations. Similarly to British Midland 30 years earlier, they mistakenly identified the operating engine as the damaged one and shut it down.
All this happened within the first 28 seconds of the incident. A few seconds later, the “L ENG FLAMEOUT” message finally appeared, and as they had the “R ENG SHUTDOWN” already, they found themselves with a “DUAL ENGINE OUT” situation, with all the associated cascading messages, emergency power, Ram Air Turbine noise, loss of screens, loss navigation, etc…
Talk about stress and sensory overload. Air France 447 was another example of such chaos in the cockpit.
Only one thing was left for the crew, divert to the nearest suitable airport while trying an engine restart but, like Pinnacle Flight 3701, they priorities the engine restart over the landing site and chose to return to homebase.
The restart never occurred because all along they were attempting to mistakenly restart the failed left engine.
More importantly, at the time of the failure, they had Kabul airport below them and a second airport within gliding distance. When they realised they would not make it, it was too late and the only option was a forced landing, ending in a tragic loss of life.
History never repeats itself, but tragically rhymes.
You can read the full report at here