Aircraft Accidents and Lessons Unlearned LVII: Kenya Airways flight KQ431

Kenya Airways, flight KQ431

On January 30, 2000, at 9:09 PM (21:09) Greenwich Mean Time (GMT) Kenya Airways flight KQ431, an Airbus A310-304, registration 5Y-BEN, with General Electric (GE) CF6-80C2A2 engines, impacted the North Atlantic Ocean 1.5 nautical miles south of Port Bouët Airport (ABJ) in Abidjan, Côte d’Ivoire, where it had departed from 33 seconds prior; Abidjan, Côte d’Ivoire is in the GMT time zone. KQ431 was the first flight leg through Lagos, Nigeria (LOS), terminating in Nairobi, Kenya (NBO). The accident occurred over two hours after sunset, 6:39 PM (18:39) GMT; the Moon, which rose four hours later, was in a waning crescent, with 25% illumination. The meteorological conditions: eight kilometers (five miles) visibility; clear with scattered clouds at 390 meters (1280 feet) and the ocean leading to the point of impact was dark against a dark sky, giving the flight crew no exterior point of reference of the horizon. Instruments were found not to be an issue.

Per accident report 5y-n000130, the Ivorian Commission of Inquiry (ICI) concluded, “… that the cause of the accident to flight KQ 431 on 30 January 2000 was a collision with the sea that resulted from the pilot flying applying one part of the procedure, by pushing forward on the control column to stop the stick shaker, following the initiation of a stall warning on rotation, while the airplane was not in a true stall situation.

To be clear, per the report’s cockpit transcript, the accident occurred when a stick shaker – a warning that alerts the pilots to a stall situation – activated nine seconds after rotation – not on rotation. The flight was committed, the gears were retracting and the aircraft was nose up in climb meaning; this meant engines were at takeoff power. Per the report’s conclusion, the first officer (FO) [pilot flying] did not complete required Flight Crew Operating Manual procedures during the stall event. Per the report, though he lowered the nose to escape the alleged stall, he failed to advance the throttles to the ‘Takeoff-Go Around’ position to increase thrust, despite the fact the aircraft was still taking off and in CLIMB.

There were too many concerns in this investigation report to speak to in one article, but three problems that stood out were:

  • Who had investigated the accident?
  • How the Maintenance investigation was conducted.
  • The captain’s actions (or inactions).

Who had investigated the accident? Per report 5y-n000130, “In accordance with Annex 13 of Article 26 of the Chicago Convention, the Ivory Coast, the State of Occurrence, launched an investigation.” The report stated, “… accident notifications were sent, in accordance with the provisions of Annex 13, to Kenya and France, respectively the State of registration and the State of Manufacture, as well as to the International Civil Aviation Organisation (ICAO). The investigation work itself began on Monday 31 January 2000 with the establishment of several working groups made up of Ivorian, Kenyan and French investigators. At the Ivorian authorities’ request, the French Bureau d’Enquêtes et d’Analyses pour la sécurité de l’aviation civile (BEA) were deeply involved in the investigation and created the report.

It was unusual that, per the accident report, Airbus, GE and Kenya Airways were not listed as providing on-site representation. Organizations listed as contacted and represented, included the BEA, the French government’s investigatory agency, as the State of Manufacture. If Airbus, GE and Kenya Airways were represented, was there concern about the level of expertise on the ICI’s part? Was the ICI up to the challenge of a major accident investigation? How many major accident investigations had the ICI overseen in the past? Would their decisions depend on the integrity of the manufacturers?

Questions of design faults or inherent safety issues that cannot be investigated with objectivity meant that the causal factors would have been lost. If the ICI relied too heavily on Airbus, GE and Kenya Airways for answers, either they might have missed any questionable issues out of ignorance, or they may have overly trusted Airbus, GE and Kenya Airways to be candid with sensitive or proprietary information that aided the investigation. All organizations are self-protective, that was why objective team leaders with experience were important to the impartial flow of information and data research.

How the Maintenance investigation was conducted. Per the report, the investigations were conducted in the following titled areas: Operational aspects, Air Traffic aspects, Aircraft, Site and Wreckage, Readout of Flight Recorders, Testimony, Meteorological aspects and Medical/Pathological aspects. It is presumed – though unknown – that ‘Aircraft’ meant Systems, Powerplants, Structures and Maintenance. Questions raised by KQ431’s unusual last minutes required that the ‘Aircraft’ be investigated by each specialty. While the Causes of the Accident focused on alleged pilot error and their purported failure to follow procedures, what the investigation did not properly convey was why the airplane had a stall warning on climb-out to begin with. The maintenance investigation into the accident was so undeveloped, it bordered on non-existent. Why did the stall warning go off on a routine flight? Was the angle of attack vane stuck? What did the attitude instruments say? Were they cross-checked with the standby indicators?

The investigators concluded that the inadvertent activation of the stall warning was not due to known possible causes, e.g., uncommanded spoiler or thrust reverser deployments; center-of-gravity issues; slat retraction out of sequence; improper speed indication; insufficient thrust per aircraft attitude or aircraft improperly configured. In the report, a list of past events where false stall warnings occurred without a traceable reason was included in Appendix 16 in French; it referred to several different airliners, which made the subjective list inapplicable.

Instead, the ICI referred to a ‘selected scenario’ to explain events that led up to the accident. How was this scenario arrived at? Why did the ICI defer to it? This is where the ICI’s – or any investigatory group’s – aircraft knowledge and experience was vital because a complete reliance on the manufacturer to be candid meant the ICI report’s quality solely depended on the manufacturer holding itself accountable. It was not about blame; it was about cause.

The report stated, “The Commission of Inquiry concluded that this malfunction [stall warning activation] existed before the last departure from Nairobi of flight KQ 430 but that the inoperative condition of the FDR had no connection with the accident.” The ICI made this statement without any basis in evidence. How did they arrive at this conclusion? Did the Maintenance investigation find any reason for this? Systems? Was there anything in the maintenance log to suggest this happened?

Nowhere in the Causes of the Accident or the Contributing Elements were any aircraft or engine concerns addressed, questioned nor even raised. Instead, the three Contributing Elements bullets wandered off on a tangent, calling out trivial issues that did nothing to find the cause. All focus was on pilot, pilot, pilot, bells and whistles – not aircraft. This attention to pilot error raises the third problem:

The captain’s actions (or inactions). A major investigation’s greatest myth is that the cockpit voice recorder (CVR) is effective, that it is productive and should be depended on for accuracy. A CVR is a tool. To unqualified investigators it is a go-to device; a crutch for the inexperienced; a distraction from other evidence that confuses more scenarios while answering less questions. The only thing more unwise would be to install cockpit cameras. But then, where in the cockpit would they be installed?

The CVR’s uselessness was noted by the report when referring to the different flight crew responses that could not be determined in KQ431. Between the first sounding of the stall aural warning until impact, 27 seconds passed. Were the captain’s and FO’s instruments giving conflicting information? Those listening to the CVR could hear aural alarms, but few words spoken; no sense of the speakers’ urgency, no recognition of what the captain was doing. In those 27 seconds, why did the captain not say, “My airplane,” then take control? Was the captain busy with the gear and flaps?

It is hard to believe, with the cacophony of aural warnings, the captain was unaware of an unfolding emergency. He had 1664 hours in the A310 while the FO had 5768 hours. Even though the FO was qualified, hours-wise, and was the pilot flying, the captain was in command.

The report itself may have inadequately portrayed the investigation. Perhaps some things were missed in translation from French to English. However, the conclusions were specific, the findings ambiguous. It is difficult for the interpretations to have missed important information when the conclusions were focused on pilot error and little else.

It is said that a pilot’s greatest friend is altitude; the higher an airplane is above the ground, the more time and room the pilot has to recover. The true frustration with KQ431 was that pilot error was never proven to be a factor; even though the captain did not assume control, the FO had checked out on the A310 and had the flight hours to prove his ability. More concerning, an Airbus A310 takes off on a routine flight before it plummets into the sea in under 30 seconds and the aircraft and its maintenance received minimal attention. Where was the industry on this? Did anyone reading the report ever ask the obvious questions?