I was talking to someone about the Northwest Flight 188 – the one where both pilots somehow managed to get so distracted that they missed their destination and did not respond for over an hour. I expressed some indignation that the FAA had revoked the licences of both pilots without waiting for the NTSB report. She wondered what difference that might make – which I found somewhat difficult to explain.

The strength of accident investigations is that it is not simply a blame-allocation exercise. 2008 NTSB statistics for scheduled flights show one accident per million flight hours, no fatalities. One of the reasons that aviation is relatively safe is because every accident is treated seriously, rather than dismissed the moment someone is found who could be held accountable.

The focus over the last few decades has been on how to avoid the same scenario or sequence of events in the future, which is critical.

A very good example of how this works is Canada’s Transportation Safety Board of the Swissair Flight 111 in 1998.

BBC News | 8 September 1998 | Flight 111’s black hole

The first that air traffic controllers heard of the problem was 16 minutes before the crash.

The pilot announced: “Swissair 111 is declaring pan pan pan – we have smoke in the cockpit.”

“Pan pan pan” is the expression used when an emergency is less acute than a mayday signal, which indicates imminent disaster. But the situation rapidly deteriorated.

The pilot suggested landing at Boston, but was told Halifax was closer, so he began heading in that direction.

However, the plane was at an altitude of around 10,000 metres and needed to lose height.

Air traffic controllers also gave the pilot permission to dump at least 30 tons of fuel to land safely to help it land safely.

The pilot’s next words on the radio were that he was declaring an emergency.

“We have to land immediately,” the pilot said, the last words the controller heard from the plane.

Radar signals showed that the airliner began flying off course in a rapidly descending loop over the sea.

Six minutes later, it hit the water.

The easy assumption was that the flight crew were remiss in not declaring an emergency immediately, wasting valuable time.

The flight crew realised there was smoke in the cockpit. They presumed it was a problem with the air conditioning. They looked up the nearest airport with appropriate mechanics and asked for a diversion. The captain went through his checklists of how to respond to smoke in the cockpit. Long story short, by the time they declared an emergency, the captain was out of his seat trying to put out a raging fire, the First Officer was desperately trying to regain control of the aircraft with all the displays failed / failing and the plane crashed into the ocean.

Again, it would be easy to blame the Captain and his first officer: why didn’t they immediately declare an emergency? Why didn’t they immediately focus on getting the plane on the ground and *then* work out where the smoke was coming from?

The investigation cost millions and the Transportation Safety Board of Canada final report wasn’t released until five years later but it identified eleven “causes and contributing factors” which basically showed that the plane itself was unsafe – the initial problem (an unsafe in-flight entertainment system) didn’t trip circuit breakers and the insulation of the plane was flammable. The immediate effects of the investigation included removing flammable insulation material from aircraft.

The investigation concluded that, even if the crew had been immediately aware of the extent of the problem, they never had enough time to save the flight. A small electrical fire lit the insulation and the fire was out of control before the plane could have ever made it to the ground, even if the crew had diverted immediately.

Many changes were made, including better electrical systems and also usage of inflammable materials in planes. However, the report also isolated a specific issue with cockpit training, specifically challenging a presumption that pilots should operate from a best-case scenario.

Pilots live by their checklists which are meant to ensure that best procedures are followed both on the ground and in the air. The flight crew are trained to carry out checks and operations in a specific order, which should be both the safest and most efficient response to the situation. The concept of a checklist has been proven effective but there were some questionable assumptions made in some checklists, especially the standard checklists for Unidentified Smoke on Board.

Most instances of smoke in the plane are benign and it is rare to have an in-flight fire. The training was focused on finding the source and eliminating the problem, rather than treating the situation as a potential serious fire threat until you’ve identified the issue. In most instances, the situation is quickly dealt with in the air but it is incredibly dangerous to presume that will be the case.

Since this accident, the discovery of smoke is treated as a potential in-flight fire (that is, an emergency) until the source is found and the problem eliminated.

The knee-jerk response was “there was an electrical fault and the pilots failed to handle the emergency” whereas the drawn-out and expensive investigation effectively proved that there was a lot more to it than that.

As a result, specific changes were recommended (and taken up) which have helped us to avoid this sequence of events in the future. These changes in themselves would not have saved Swissair Flight 111 but their impact on aviation safety for the future is inestimable.

For more information, read the full accident report: Transportation Safety Board of Canada – AVIATION REPORTS – 1998 – A98H0003

If you found this post interesting you might enjoy the following:

• Air France 447 Update on Investigation
• 15 Years since ValuJet Flight 592
• Human Factors: Crossair Flight 850
• Biggin Hill Accident Report
• Alaska Airlines Flight 1866
• PULL UP PULL UP: Garuda Indonesia flight GA200 in Yogyakarta