- Cause Mapping®
- Tools & Resources
- About Us
Join us for the next Cause Mapping Root Cause Analysis Public Workshop ONLINE on June 13-15!
On the morning of July 25th, 2000, passengers boarded Air France Flight 4590 from Paris to New York and settled in for what was supposed to be a long flight on a supersonic aircraft. Sadly, their flight lasted less than two minutes. Just after liftoff, the supersonic jet crashed into a hotel in Gonesse, France, killing all 109 people aboard and an additional 4 people on the ground.
Five minutes before Flight 4590 took to the runway, a Continental flight headed to Newark, using the same runway, lost a titanium alloy strip. Normal protocol for a Concorde flight includes a full runway inspection before takeoff; this was not completed (perhaps because the flight was already delayed by an hour). During Flight 4590’s takeoff, a piece of this debris from the Continental flight, cut and ruptured one of the Concorde’s left tires. As the aircraft accelerated down runway 26R, this tire disintegrated and a piece of it struck the underside of the wing, where fuel tank 5 was located.
A pressure wave inside the tank caused it to rupture forward of the tire strike. Fuel poured from the tank and ignited. The Concorde had already reached a velocity where it could not stop safely by the end of the runway so it lifted off the runway with flames hanging from the left wing. There are some incredible photos that captured this amazing moment, a moment that cost the lives of 113 people, $125 million, and the heretofore stellar reputation of a truly impressive airliner.
The history of supersonic commercial air travel has its roots in the 1950s and 1960s, the same period that witnessed the Cold War American and Soviet spaceflight rivalry that launched a man into space. While the Cold War superpowers jockeyed to conquer the stars, Britain and France set its eyes on the skies with the ambition of air travel faster than the speed of sound–faster than commercial flights had ever flown–a reality.
In the late ‘50s, the US, France, Britain, and the Soviet Union all toyed with the idea of supersonic flights. British and French companies, in large part funded by their governments, developed designs that were ready to go to construction by the early 60s, but the cost of such an ambitious project proved too prohibitive for either to accomplish alone. In 1961, therefore, British Aerospace and France’s Aerospatiale came together to produce and develop the project, whose development was negotiated not as a commercial agreement between the respective companies, but as an international treaty between the nations; the treaty was signed in 1962. The Concorde takes its name from this agreement (the word concorde in French (concord in English) means “agreement, harmony, or union”).
Two prototypes began construction in 1965 and were presented to the public in 1969, the same year as their first test flight, at the Paris Airshow. In January 1976, Concorde celebrated its first commercial flight.
As impressive as this accomplishment was on its own, for the French and the British it had always stood for something greater. The Concorde was a symbol of national pride for the post-Imperial countries, a way of remaining on par with the two superpowers that emerged from the Second World War for the two nations that came out of that conflict in desperate search of new sources of identity in a world that had left their greatest glories. Both nations, but especially France, were in need of a source of national pride. Britain’s “finest hour” of bravery during the war was fading into history, and its empire was crumbling. France, for its part, had been humiliated and occupied during the war.
When flight F-4590 crashed near Paris, then, it was a huge blow to the brand as well as to the nations that had nurtured it.
In discussing incidents of this magnitude that involve not only loss of life but also extremely sophisticated and complex technology, it is often difficult to break down an incident to the extent that anyone–not just engineers or people who work in the industry–can understand them. Root cause analysis is a powerful tool for doing so. As we have demonstrated in our coverage of similar incidents in spaceflight, the Cause Mapping approach to root cause analysis analyzes incidents in terms of a detailed chain of cause and effect, promoting a better general understanding of the event at hand as well as multiple opportunities to enact solutions that prevent such catastrophes from happening again.
To get a better idea of what caused the crash of the Concorde, we built our very own cause map- ThinkReliability’s root cause analysis tool- to order the causes and tie them to organizational goals. Cause Maps break incidents down into their individual contributing elements and are thus fantastic tools for understanding disasters like these (see space maps) fully and deeply.
Our snapshot of the Concorde incident begins with a detailed picture of the incident–what happened–before we get into asking why. Below you will find our problem summary for the Concorde disaster, including the problem, the time and location, and the industry goals that were affected by the event.
Approaching an incident from a goals perspective solves several issues that one may otherwise encounter in undressing an incident. First, it prevents a root cause analysis free-for-all that fails to stick to the roots of the incident. Focusing on problems leads to blame and arguments rather than to proactive problem-solving.
Focusing on problems also removes the team performing the root cause analysis from the key purpose of the discussion in the first place: to make for a better future for the organization and all involved.
So, instead of asking what went wrong (this can elicit a number of answers, some more productive than others), we start with our goals as they relate to the ideal state of the airline. Naturally, fatalities are bad. Resource loss is also bad, but most people would agree that an organization that loses money but not lives is closer to its ideal state than one that saves money but loses lives.
The number of fatalities- 113- is broken into two separate causes of a safety goal incident because it is useful for us to remember that some were killed because they were on the plane and some were killed because they were in the hotel.
Sometimes, asking ourselves “Why?” over and over again can produce great results when it comes to pinpointing and solving problems. The 5 Whys Technique is a way to kickstart any root cause analysis because it immediately creates building blocks for your cause map.
Why did we have fatalities? Because the Concorde crashed.
Of course, reviewing the “no crashing planes” policy won’t solve our problem. So we ask “Why?” four more times and wind up with something like this:
The 5 Whys Technique is clean and effective because it allows you to begin building out the cause map immediately.
By using this technique in our root cause analysis, we find that the Concorde’s crash was caused by the loss of the two engines on the left side of the aircraft. A piece of debris on the runway caused one of the left side tires to disintegrate. When the tires exploded a piece hit the underside of the aircraft, which ruptured one of the fuel cells slightly ahead of the intakes to the engines 1 and 2. The fuel, which ignited, choked out the two engines on the left side, and the Concorde crashed into a hotel in Gonesse, France just 5 km from the runway.
This approach to root cause analysis is an excellent way to approach any issue because of the basic structure it provides. Continuing with our root cause analysis, we can add cause and effect relationships to the left or right, in between, or vertically from each of the causes on the cause map:
Air France Flight 4590 was the Concord’s only fatal accident in its 31-year history (no other commercial aircraft has matched that record). At the time, with a record of zero accidents per km traveled before the accident, the Concorde qualified as the safest airliner in the world. Nonetheless, the crash of Air France Flight 4590 marked the beginning of the end for the mythic airliner.
Both Continental Airlines and John Taylor, one of its mechanics, were found criminally responsible for their part in the disaster in December 2010, but their convictions were overturned in a French court in 2012, on the grounds that the mistakes that they made did not amount to criminal responsibility.
The first passenger flight in the wake of the accident took to the skies on September 11, 2001, and touched down just before the world trade center attacks for which that date is better remembered.
Air France and British Airways announced the retirement of the Concorde fleet in April of 2003, citing lower passenger numbers after the Air France crash, compounded by the general slump in air travel that followed 9/11 and increased maintenance costs.
Airlines are generally thought of as high-reliability organizations because there are over 87,000 flights per day in the United States alone (according to the National Air Traffic Controllers Association), yet we have not seen a fatal major airline crash in the United States since November of 2001 when an American Airlines flight A300 crashed into a residential neighborhood in Queens, New York, not far from the departing airport. Killed 5 on the ground, all 9 crew members and 251 passengers, and damaged several homes. In that incident, the airplane had broken up while in flight–having shed its vertical stabilizer, left rudder, and engines, the plane spiraled out of control and landed on a house.
Though such incidents are by their nature dramatic and thus widely reported when they occur, aviation is generally a high reliability–a safe–industry. Of the of the 34,434 transportation fatalities in 2011, only 494 related to aviation.
Part of the key to high reliability is following procedure- every time. Sometimes, when an organization goes long enough without an incident, members of the organization may begin to question procedure:
“We know this stuff. Do we really have to go through our checklist every time?”
The answer from a root cause analysis perspective?
Yes. Every time.
Without a checklist, an organization begins to lack order. Like in much of what we do, it’s the little things that count.
High-reliability organizations rarely undergo root cause analysis because they rarely need to. Instead of assigning blame to individuals rather than the organization, these organizations consider the incident in the context of organizational goals. That way, root cause analysis can focus on solutions rather than blame. Blame can actually reduce accountability, the same way checklists can increase it.
As in all other aspects of life, root cause analysis is a place where we constantly encounter uncertainty. When we come across uncertainty while building a cause map, we simply use a question mark or mark a causality fork with “OR” to indicate a “this or that” relationship between two causes.
When we have all of the necessary evidence, this becomes unnecessary. Uncertainty only matters when we are lacking evidence. For most past incidents, there is an ample amount of evidence to make important determinations.
The complexity of the cause map we build is based on what would be useful to us. In the case of the Concorde crash, we don’t need to examine why the hotel was built where it was, or why the patrons who were unfortunately killed were standing where they were at the time of the incident.
Rather, it would serve us better to acknowledge that there is often uncertainty, unknowable purposes behind things, and many of them may not contribute to the root cause analysis itself because it won’t explain the incident in a useful way.
Because the Cause Mapping approach to root cause analysis draws its power by virtue of being so visually appealing and easy to understand, we have produced a poster of the Concorde Cause Map, containing a basic and detailed analysis of the accident. While the basic Cause Map starts with five Why questions, the detailed analysis of the Concorde accident contains over 100 cause-and-effect relationships. The solutions from the actual investigation are located above the particular cause it controls and are additionally summarized in a numbered table. To order your own Concorde incident poster, click here.
Schedule a workshop at your location to train your team on how to lead, facilitate, and participate in a root cause analysis investigation.