It is interesting reading the full report in the light of the MCAS involved crashes, especially the Lion Air crash.
The
Dutch report on the crash of Turkish TC-JGE, 25 Feb 2009 made a point of asking Boeing about the procedure when an instrument is identified as inoperative during the flight (Section 1.2.4 on page 18), but in Appendix N, Similar Occurrences (p 204)
TC-JGE was involved in two similar occurrences, which became known by the flight data recorder.
Incident A
TC-JGE made an ILS approach for runway 27L with two autopilots engaged at London Heathrow airport in England on 23 February 2009. The captain’s radio altimeter system (left) displayed a negative value and both autopilots disengaged when an altitude of 500 feet was passed. The auto-throttle ‘retard flare’ mode was activated and at this point the throttles moved aft. The airspeed dropped below the selected speed. After four seconds the crew disengaged the autothrottle and manually brought the throttles forward. Subsequently, the aircraft landed safely.
Incident B
TC-JGE made an ILS approach for runway 23R with the right autopilot engaged at Damascus air-port in Syria on 24 February 2009. The captain’s radio altimeter system (left) specified a negative value when an altitude of 4000 feet was passed. After the flaps were selected at approximately 2500 feet, the autothrottle activated the ‘retard flare’ mode. The airspeed at this time was 209 knots and the selected speed was set on 155 knots. After having flown at 1500 feet for 74 seconds, the airspeed went below the selected speed. The speed was 16 knots below the selected speed at 1400 feet. The throttles were moved forward; a nose up movement ensued with an increase in alti-tude and speed. Subsequently, the crew reduced the selected speed to 138 knots. A few seconds later the crew disengaged the autothrottle and autopilot. The ‘retard flare’ mode had been activate for 94 seconds.
So, much like the problems associated with the AoA indicator in Lion Air, there was a history of problems with the RA, and the third pilot in the cockpit even mentioned the RA fault on the crash flight - but in this case, the flight continued to have problems that were not solved in time to prevent the crash.
Like many crashes, there are a lot of holes leading up to the final incident. I do not wish to fan the 'pilot error' flames by providing an example where two previous crews resolved the issues successfully and thus saying the third set were deficient: rather this seems to point to a systemic fault that was not identified and relied on humans to solve the problem: which, as we know, will not work every time.
It is also worth underlining that the pilots on the fatal crash did not know that the autothrottle would continue to rely on the faulty radio-altimeter. The information was available to Boeing, and published in documentation not available to the pilots, as described by
Decker's report on page 36.
The human factors failures here are glaring.