Mayday at 41,000 Feet—Watch Those Units!

As a long-time science teacher, many times I get answers like this to mathematical problems: "Twelve."

Like any well-trained dog—I mean, teacher—I automatically blurt back, "Twelve what? Are we talking about 12 eggs, 12 pencils, 12 pounds, 12 liters, or 12 pieces of metababbaboobinite? Twelve what?"

Watching for mathematical units is one of those lessons that students just never seem to want to pay attention to. Homework upon homework, test upon test, my students consistently leave the units off their work.

Now, I am sure that you are familiar with the case of NASA's Mars Climate Orbiter, which crashed into the surface of the red planet back in September 1999. Within weeks, an investigation board determined that the NASA engineers had failed to convert pounds to newtons when measuring rocket force. The monetary cost to the people of the United States and the embarrassment to NASA were hefty. Yet, no human life was ever placed in danger.

On Saturday, July 23, 1983, Air Canada Flight 143 was to suffer a somewhat similar fate. At 132 tons, the three-month-old twin-engine Boeing 767 was a behemoth. To the casual observer, its cockpit glowed like it belonged in a video arcade. Loaded with screen upon screen of state-of-the-art computerized instruments, this was a plane that could seemingly fly itself. The most sophisticated jetliner of its time, it was the pride of the Canadian fleet.

Flight 143 departed from Montreal destined for Edmonton, with a brief stopover in Ottawa. The plane was certainly in very good hands. Captain Robert Owen Pearson was a 26-year veteran with Air Canada. For years, Pearson had combined his career as a commercial pilot with work as a glider instructor and tow-plane operator. He and First Officer Maurice Quintal were among the few elite at the time who were trained to operate this incredible machine.

Along the final leg of the trip, the plane was flying at an altitude of 41,000 feet and everything seemed to be going just fine. Out of the blue, at 0109 Greenwich mean time, a warning light and one of those oh-so-annoying buzzer alarms suggested a problem with the forward pump in the left fuel tank. With two pumps in each of the plane's three fuel tanks, there was little cause for alarm. Redundancy was built into every aspect of the plane's operation. Then, suddenly, alarms went off indicating a problem with the second pump in the left-wing tank.

Uh, oh!

Without hesitation, Pearson contacted the Winnipeg Air Traffic Control Center with word of a problem. Flight 143 was immediately given clearance to land at Winnipeg, which was about 130 miles south of their location at that moment.

Things were only to get worse. Much worse.

Within five minutes of the first warning, alarms indicated that all six of the fuel pumps on the plane were failing. At nine minutes, the left engine failed. At 12 minutes, the right engine failed and the plane became powerless. The cockpit fell into total darkness and auxiliary power kicked in, only to die out shortly thereafter.

This was the worst possible scenario. Both engines had failed, the auxiliary power system had given out, and the plane was a long way from its destination. This was certainly one plane that was never designed to fly without power. While it wasn't falling out of the sky, it was pretty obvious that this plane was not going to stay up forever. The only thing that seemed to work in their favor was something called the ram air turbine, which used wind power to generate minimal hydraulic control. Clearly, Flight 143 had run out of fuel.

Calculations both within the cockpit and on the ground confirmed that without power, this plane was not going to make the Winnipeg airport. Instead, the decision was made to make an emergency landing at Gimli in Manitoba, an abandoned Royal Canadian Air Force base that had been equipped with twin 6,800-foot runways. Landing at Gimli did pose some additional risks, since the base lacked both a control tower and emergency equipment.

Would they make it? Could they make it? Even if they did, could the plane be controlled with enough precision to keep it from smashing into the ground or rolling over upon impact? No one knew the answers to these questions, but when you are faced with either making a very risky landing or death itself, the choice of what to do is very clear.

As the plane approached the Gimli airfield, the landing gear was released. Without any power to assist, the pilots were dependent on the force of gravity alone to lock the landing gear in place. The main landing gear fell into place without any problem. They weren't as lucky, however, with the nose gear. The onslaught of the wind beneath the plane prevented the front landing gear from locking into place. Now flying a powerless giant, the pilots were faced with the reality of landing it without all of the landing gear down.

If you thought that things couldn't get worse, just read on....

The plane was coming in too high and risked overshooting the runway. In a move that was unheard of for a large aircraft, one that he had picked up from his gliding experiences, Pearson put the plane into a maneuver known as a sideslip. Although frightening to all aboard, it caused the plane to slow down and lose altitude, and Pearson brought the plane down within 200 feet of what would be considered a perfect landing.

The plane was down, but it had not stopped. Almost immediately, two tires in the right main landing gear popped and the casing of the right engine was scraping the pavement. Without the front nose gear in place, the belly of the plane was soon doing the same. Sparks were flying, but the increased force of friction actually helped slow the plane down sooner.

*  *  *

There was also another unexpected problem. There were people in the middle of the runway! Unbeknownst to the pilots, the former runway was now the final straightaway of the two-kilometer Winnipeg Sports Car Club racecourse. Yes, you read it correctly. They had chosen to land on an active racetrack. People were scrambling for their lives.

All of a sudden, Pearson noticed a metal guardrail, which had been put in place for drag races, running right down the middle of the runway. He tried to veer the plane to the right side, but had no luck. The left side of the plane's nose sheared the posts of the guardrail right off at their base. The plane finally came to a stop in a huge cloud of smoke. Amazingly, injuries from this near disaster were minimal and mostly caused during the evacuation of the plane.

So how did the plane run out of fuel?

Just like with my students, the error was basically in the math. Prior to the flight, the plane had experienced problems with its Fuel Quantity Indicating System, which basically controls the entire fueling process and all of its onboard fuel gauges. Generally, all the ground crew needs to do is to hook up the hoses and dial in how much fuel is needed. The computer does the rest. Since the plane was new, no spare parts were available to replace the faulty Fuel Quantity Processing Unit. The ground crew had to resort to calculating the fuel load by hand. They used the tried-and-true fuel-drip procedure, which is very similar to the dipstick that you have in your car to measure the amount of oil in the crankcase.

This manual accounting of the fuel led to another problem. The flight plan called for 22,300 kilograms of fuel, but the fuel trucks measured their fuel in liters. A multiplier of 1.77 was used to convert between the two units, but it was later learned that this was the wrong conversion factor. Instead of converting the liters to kilograms, they had calculated the number of imperial pounds. Since these were the first planes in the entire Air Canada fleet to use metric measurements, not a single person involved realized that the wrong multiplier was being used. Even though the drip test was repeated several times by different people, the conversion factor was so familiar that no one questioned its value. It was a number that they had all used over and over again in their calculations. This error, which was compounded by poor communication and training, allowed the plane to fly off with approximately half of the fuel that it actually needed to make its destination, and it almost proved deadly.

They say that history has a way of repeating itself, and this story is no exception. Eleven months later to the day, on June 23, 1984, Captain Pearson and First Officer Quintal were once again behind the controls of the same exact plane on its way from Ottawa to Montreal. During takeoff, the same sequence of lights and buzzers indicated problems with the fuel pumps. When they leveled off, everything returned to normal. This time they were not taking any chances and immediately returned to the ground. It was later determined that they were in no danger. This time the plane had not run out of fuel. Instead, a false alarm had been generated by the plane's rapid ascent. (We can be sure that these guys were not happy campers....)

The moral of this story? Watch those units! Failure to do so could have devastating consequences.

Steve Silverman teaches physics, earth science, and computer science at Chatham High School in Chatham, New York. He hopes that his love of researching, writing, and sharing bizarre true stories will show students that learning is enjoyable. To read more of Silverman's stories, look for his books, Einstein's Refrigerator and Lindbergh's Artificial Heart, and visit his Web site, www.uselessinformation.org. This article is excerpted with permission from Lindberg's Artificial Heart, Andrews McMeel Publishing: Kansas City, 2003.