The 767 That Flew Like a Glider
At 41,000 feet over Canada, a brand-new jetliner carrying 69 people lost both engines at once. What happened next depended on a math error nobody caught, a pilot's forgotten hobby, and a runway that wasn't supposed to be a runway anymore.
On the afternoon of July 23, 1983, Air Canada Flight 143 was cruising at 41,000 feet somewhere over northern Ontario, halfway through a routine hop from Montreal to Edmonton with a stop in Ottawa. It was a brand-new Boeing 767, one of the most advanced airliners in the world, carrying 61 passengers and 8 crew. Then a warning chime sounded in the cockpit, and a light came on indicating low fuel pressure in the left engine.
It shouldn’t have been possible. The instruments said there was plenty of fuel left. But the instruments were wrong, and within minutes both of the plane’s engines would be dead, leaving a 132-ton aircraft with no power, at cruising altitude, over a stretch of Canada with no obvious place to land. What happened over the next twenty minutes has become one of the most studied incidents in aviation history — not because of what went wrong, but because of how it very nearly didn’t go wrong at all, and then, at the last possible moment, went wrong again in a completely different way.
A new kind of airplane, and an old kind of scale
To understand how a functioning airliner could simply run out of fuel, you have to start with a unit conversion.
Air Canada’s fleet, like the rest of the aviation industry in North America at the time, measured fuel the old way: in pounds, calculated from a density factor of 1.77 pounds per litre. But Canada was in the middle of converting to the metric system, and the Boeing 767 that became Flight 143’s aircraft, tail number C-GAUN, was the first aircraft in the Air Canada fleet built to work in metric units — kilograms and litres, with a fuel density factor of 0.8 kilograms per litre. It was a new airplane with a new cockpit built around a new national standard, delivered into an airline whose ground crews, dispatchers, and maintenance systems had not yet fully made the switch.
That mismatch alone might have been manageable. What turned it into a crisis was a broken sensor. The 767’s fuel quantity indicator system — the automated system meant to tell the crew and ground staff exactly how much fuel was in the tanks — was malfunctioning on the flight before Flight 143, and had been for days. Rather than ground the aircraft, mechanics in Edmonton disabled the faulty part of the system and told the crew to keep flying, using a manual backup procedure: a mechanic would dip a physical measuring stick, called a “drip stick,” into the tanks to read the fuel level directly, then hand-calculate the volume of fuel to load using the conversion factor.
In Montreal, before the flight to Ottawa and on to Edmonton, that manual calculation was performed twice — once by the maintenance crew and once again, as a check, by the flight crew themselves. Both times, the person doing the math used 1.77, the old pounds-per-litre factor, instead of 0.8, the correct kilograms-per-litre factor for this aircraft. The error meant that instead of loading the roughly 22,300 kilograms of fuel the flight needed, ground crews loaded barely half that amount. Every instrument in the cockpit, including the computer that was supposed to catch exactly this kind of mistake, reported the numbers as correct, because the numbers had been calculated correctly according to the wrong formula. Nobody — not the mechanics, not the two pilots who cross-checked the math, not the dispatch system — caught it.
Two flameouts and a checklist that didn’t exist
Captain Robert Pearson, 48 years old, had more than 15,000 hours of flying experience. His first officer, Maurice Quintal, had flown for the Canadian Armed Forces before joining Air Canada. Neither of them had any reason to doubt the fuel numbers they’d been given in Montreal — the math had checked out on paper.
The first indication that something was deeply wrong came with the fuel pressure warning for the left engine, which both pilots initially treated as a possible sensor fault, consistent with the aircraft’s already-known instrumentation problems. Then the left engine flamed out entirely. While they were diverting toward Winnipeg to deal with the single failed engine, the second warning came, and moments later the right engine died too. The cockpit went eerily quiet. The 767’s main electrical systems, powered by the engines, dropped offline, and most of the cockpit instrument panel went dark. A small emergency propeller called a ram air turbine — designed to drop automatically out of the belly of the aircraft and spin in the airstream — deployed and began generating just enough hydraulic and electrical power to keep the flight controls and a minimal set of instruments working.
Pearson and Quintal searched the emergency checklist for a procedure covering a dual-engine failure at altitude in a 767. There wasn’t one. It wasn’t considered a survivable scenario worth writing a checklist for, because a fully loaded, modern twin-engine jet losing both engines simultaneously from fuel exhaustion was thought to be so unlikely that no one had planned for it.
What they did have was distance, altitude, and two pieces of unrelated personal history that turned out to matter enormously.
Silence in the cabin
Behind the cockpit door, the passengers had no idea about litres, drip sticks, or conversion factors. What they knew was that the engine noise had stopped, the cabin lights had dimmed, and the flight attendants — who had been trained never to let their own fear show — were moving through the aisles calmly demonstrating the brace position and pointing out emergency exits as if this were simply a routine safety briefing running unusually late. By multiple accounts from people on board, the cabin fell eerily quiet: no engines, no violent shaking, no smoke, just a large aircraft gliding, silently, toward the ground. Some passengers, not fully grasping how serious the situation was, stayed calm simply because nothing outwardly dramatic seemed to be happening. Others understood exactly what an unpowered airliner losing altitude over the Canadian countryside meant, and used the time to scribble notes to their families on whatever paper they could find in their bags.
A glider pilot’s math and an air force base that wasn’t supposed to exist anymore
Pearson happened to be an experienced glider pilot in his personal life, flying sailplanes with no engine at all as a hobby. That background gave him an intuitive feel for how to fly a large aircraft with no thrust: he needed to find and hold the correct glide speed to travel the maximum distance for the altitude he had, and every excess knot of speed or unnecessary control movement would cost him precious kilometres of range. He and Quintal worked out a target glide speed of around 220 knots and began calculating whether they could reach any airport.
Their first idea was Winnipeg, but Quintal’s math suggested they would come up short and arrive too high and too fast even if they made it, in an aircraft with no engines to power a normal, controlled approach. Then Quintal remembered something from his own past: he had once served at a Royal Canadian Air Force base near the town of Gimli, Manitoba, and he knew its runway, decommissioned years earlier and partly converted to civilian and recreational use, might still be usable. It was closer than Winnipeg and within their glide range. Pearson turned the powerless jet toward Gimli.
What neither pilot knew was that the old air base runway was no longer just sitting empty. Part of it had been converted into a drag strip, and that Saturday happened to be “Family Day,” a day of racing and public events organized by the Winnipeg Sports Car Club. There were cars, campers, and people, including children on bicycles, on and around the runway the 767 was about to land on.
Landing without engines, without full instruments, and without warning
As Gimli came into view, Pearson realized the aircraft was coming in far too high and fast to land safely — a problem, because with no engine power there were no normal speed brakes or flaps available to slow the descent the usual way. He responded with a maneuver almost never used in commercial aviation: a forward slip, a technique common among glider and small-aircraft pilots, in which the plane is flown nose-pointed one way while banked the other, deliberately putting it into an aerodynamically inefficient, crossed-controls attitude that increases drag and steepens the descent without building up extra speed. Executing a forward slip in a widebody passenger jet was not something Boeing engineers had designed the 767 to do, and it is still cited today as one of the more remarkable pieces of stick-and-rudder flying in airline history.
On the ground at Gimli, the racers and spectators scattered as the silent, engineless jet dropped out of the sky toward the strip. The landing gear had been lowered using a gravity-drop backup system because there was no hydraulic pressure to power it normally, and the nose wheel failed to lock fully into place. When the aircraft touched down, the nose gear collapsed and the plane’s nose scraped along the runway in a shower of sparks, while the main gear tires, taking the full weight of the aircraft under heavy braking with no reverse thrust to help slow it, blew out. The aircraft screeched to a stop with its nose crumpled against the pavement, just short of the people and equipment still on part of the strip. A small fire broke out near the nose and was quickly put out by racers and course workers who ran over with hand-held fire extinguishers before the plane’s own crew could even reach it.
All 69 people aboard survived. About a dozen passengers were hurt, almost all of them minor injuries sustained sliding down the rear emergency evacuation slide, which had been left at a steep, near-vertical angle because the collapsed nose gear had tipped the whole aircraft forward. For a moment after the plane finally stopped moving, the cabin was silent again — then, according to those on board, it erupted into cheers and applause.
Blame, then a medal
In the aftermath, investigators had to untangle a chain of failures spanning the airline’s shift to metric units, a known-faulty fuel gauge that had been worked around rather than fixed, and a manual calculation checked by two separate people who both used the same wrong number. Air Canada’s own internal inquiry initially placed the blame heavily on the ground crew and flight crew, and disciplined them: Pearson was demoted for six months and Quintal was suspended for two weeks, alongside several ground maintenance staff.
A subsequent government aviation safety review took a different view, concluding that the deeper fault lay with Air Canada’s own management and its handling of a known equipment problem and an incomplete transition to metric measurement, rather than with the individuals who had been handed bad numbers and an inoperative fuel gauge. Pearson and Quintal were reinstated, and two years later, in 1985, they were both awarded the Fédération Aéronautique Internationale’s first-ever Diploma for Outstanding Airmanship, an international recognition for the flying skill that got 69 people safely on the ground.
The aircraft itself, repaired at the Gimli airfield well enough to be flown out days later, went back into full commercial service with Air Canada and quietly flew passengers for another 25 years, carrying the nickname “the Gimli Glider” for the rest of its career. It flew its final commercial flight on January 1, 2008, and was retired to an aircraft storage facility in the Mojave Desert in California later that month, with Pearson, Quintal, and several of the original flight attendants from Flight 143 on board for the trip.
The airplane that should have crashed instead became one of the most famous survival stories in aviation, built entirely out of a decimal-point mistake, a broken gauge nobody replaced, a pilot’s weekend hobby, a first officer’s years-old memory of a base he used to be stationed at, and a runway that, for one Saturday afternoon in 1983, was supposed to be a drag strip and not an emergency landing site at all.
Flight 143 is still taught in aviation schools today, not as a story about heroics alone but as a case study in how failures compound: a unit-conversion mix-up during a national metric transition, a malfunctioning instrument that was worked around instead of fixed, and a double-checked calculation that both checkers got wrong the same way. None of those failures, on their own, should have been able to bring down a modern jetliner. Together, they nearly did — and the only reason 69 people walked away from a Canadian drag strip that afternoon was that the two men in the cockpit happened to carry, entirely by chance, exactly the two pieces of unrelated knowledge the moment required.