Lead Page Photography by Wayne Glowacki

The Gimli Glider

by Wade H. Nelson

If a Boeing 767 runs out of fuel at 41,000 feet, what do you have? Answer: A one hundred and thirty two ton glider with a sink rate of over 2000 fpm and barely enough hydraulic pressure to control the ailerons, elevator, and rudder. Put veteran pilots Bob Pearson and cool-as-a-cucumber Maurice Quintal in the cockpit and you've got the unbelievable but true story of Air Canada Flight 143 -- known ever since as the Gimli Glider.

Flight 143's problems began on the ground in Montreal. A computer system called the "Fuel Quantity Information System Processor" or "FQIS" manages the 767 fuel loading process. The FQIS controls all of the aircraft's fuel pumps and drives all the gauges, leaving little for the crew and refuelers to do but hook up the hoses and dial in the desired fuel load. The FQIS was not working properly on Flight 143 due to a poorly soldered sensor. An impossibly complex sequence of mistakes by Air Canada technicians investigating the FQIS problem (which took over a year to unravel) managed to defeat the redundancy built into the system. This left Quintal and Pearson without working fuel gauges.

In order to complete their flight from Montreal to Ottawa, and on to Edmonton, Flight 143's maintenance crew resorted to manually calculating the 767's fuel load using a procedure known as "dripping" the tanks. "Dripping" might be compared to calculating the amount of oil in a car based on the dipstick reading. Among other things, the specific gravity of jet fuel is needed to make the proper drip calculations.

The flight crew had never been trained how to perform the drip calculations. It was a job previously delegated to the flight engineer, before Air Canada and other airlines went to two-man cockpits. To be safe they and others re-ran the numbers three times to be absolutely sure the refuelers hadn't made any mistakes -- each time using 1.77 pounds/liter as the specific gravity factor. This factor was written on the refueler's slip and used on all of the other planes in Air Canada's fleet. The factor the refuelers and the crew should have used on the brand new, all-metric 767 was .8 kg/liter of kerosene.

After a brief hop Flight 143 landed in Ottawa, and Pearson had the 767 re-dripped. The refuelers reporting the plane as having 11,430 liters of fuel contained in the two wing tanks. Pearson and Quintal, using the same incorrect factor used in Montreal, calculated they had 20,400 kilos of fuel on board. In fact, they left for Ottawa with only 9144 kilos, roughly half what would be needed to reach Edmonton.

Gimli from the air. Photography by Wade Nelson.
Lacking "real" fuel gauges Quintal and Pearson manually keyed 20,400 into the 767's flight management computer. The flight management computer would keep rough track of the amount of fuel remaining by subtracting the amount of fuel burned from the amount (they believed) they had started with. Their fate was now sealed.

According to Pearson, the crew and passengers had just finished dinner when the first fuel pressure light came on. Flight 143 was outbound over Red Lake Ontario at 41,000 feet and 469 knots at the time. The 767s "Engine Indicator and Crew Alerting System" (EICAS) beeped four times in quick succession, alerting them to the fuel pressure problem. "At that point" Pearson says "We believed we had a failed fuel pump in the left wing, and switched it off, and opened a crossfeed valve so the left engine could feed from the right tank. We considered the possibility we were having some kind of a computer problem. Onboard computers were pretty new back then, and we were still a bit suspicious of them. Our flight management computer showed more than adequate fuel remaining for the duration of the flight. We'd made fuel checks at two waypoints and had no other indication of a fuel shortage." When a second fuel pressure warning light came on, Pearson felt it was too much of a coincidence and made an instantaneous decision to divert to Winnipeg. Quintal keyed the mike and requested clearance in Apollo 13 fashion: "We have a problem." he said "We're going to, uh, requesting direct Winnipeg" Pearson throttled back the engines, and Flight 143 turned to the southwest and began a gradual descent to 28,000.

Skid marks in runway left by 767. Photography by Maurice Quintal.
Says Pearson, "Circumstances then began to build fairly rapidly." Pearson switched on the unused center tanks, hoping to scavenge any fuel remaining in them from a prior flight. Within moments, other fuel pressure gauges began lighting up. Nine minutes from the onset of problems, the left engine flamed out amidst a symphony of buzzers, bells, and warning lights flashing in the cockpit.

Pearson and Quintal immediately began making preparations for a one engine landing. Two minutes later, just as preparations were being completed, the EICAS issued a sharp bong - indicating the complete and total loss of both engines. Says Quintal "It's a sound that Bob and I had never heard before. It's not even in the simulator." Things got quiet. Starved of fuel, both Pratt & Whitney engines had now flamed out. Pearson's response, recorded on the cockpit voice recorder for all eternity, was "Oh F___."

767 "Glass Cockpit." Photography by Maurice Quintal
At 1:21 GMT, the forty million dollar, state-of-the-art 767 had become a 132 ton glider. The APU, designed to supply electrical and pneumatic power under emergency conditions was no help because it ran off the same fuel tanks as the engines. At 28,000 feet the 767's "Glass Cockpit" went dark. Pearson was left with a radio and standby instruments, noticeably lacking a vertical speed indicator - the glider pilot's instrument of choice. Hydraulic pressure was falling fast and the plane's controls were quickly becoming inoperative. But the engineers at Boeing had foreseen even this most unlikely of scenarios and provided one last failsafe - the RAT.

The RAT is the 767's Ram Air Turbine, a propeller driven pump tucked under the belly of the aircraft. The RAT can supply enough hydraulic pressure to move the control surfaces and enable a dead-stick landing. The loss of both engines causes the RAT to automatically drop into the airstream and begin supplying hydraulic pressure.

As Pearson began gliding the big bird, Quintal "got busy" in the manuals looking for procedures for dealing with the loss of both engines. There was nothing. Neither he or Pearson had ever been trained on this contingency.

Pearson reports thinking "I wonder how it's all going to turn out." Controllers in Winnipeg began suggesting alternate landing spots, but none of the airports suggested, including Gimli, had the emergency equipment they would need for a crash landing. The 767's transponder had "Gone dark" as well, leaving controllers in Winnipeg scrambling, using a cardboard ruler on a standby radar screen to try and determine the 767's location and rate of descent. The more modern screens wouldn't display an aircraft without a functioning transponder.

Pearson glided the 767 at 220 knots, his best guess as to the optimum airspeed. There was nothing in the QRH manual about minimum sink. The windmilling engine fans were creating enormous drag, giving the 767 a sink rate of somewhere between 2000 and 2500 fpm. Co-pilot Quintal began making glide-slope calculations to see if they'd make Winnipeg. The 767 had lost 5000 feet of altitude over the prior ten nautical (11 statute) miles, giving a glide ratio of approximately 11:1. ATC controllers and Quintal both calculated that Winnipeg was going to be too far a glide-the 767 was sinking too fast. "We're not going to make Winnipeg" he told Pearson. Pearson trusted Quintal, and immediately turned north towards Gimli.

"For me it was a cold, unemotional experience. It was the only time in my entire career I felt like I was a computer. I had a job to do and I stayed totally focused."

- Pilot Bob Pearson

Gimli, the site of an abandoned Royal Canadian Air Force Base, 12 miles away, was their last chance as a possible landing spot. It wasn't in Air Canada's equivalent of Jeppensen manuals, but Quintal was familiar with it because he'd been stationed there in the service. Unknown to him and the controllers in Winnipeg, Runway 32L (left) of Gimli's twin 6800 foot runways had become inactive, and was now used for auto racing. Several steel guard rails had been installed down the southeastern portion of 32L, dividing it into a two lane dragstrip. This was the runway Pearson would ultimately try and land on, courting tragedy of epic proportions.

To say that runway 32L was used for auto racing is perhaps an understatement. The inactive runways had been "carved up" into a variety of racing courses, including the aforementioned dragstrip. Drag races were the only racing event not taking place on April 23rd, 1983 since this was "Family Day" for the Winnipeg Sports Car Club. Go-cart races were being held on one portion of runway 32L, and just past the dragstrip a portion of the runway served as the final straightaway for a road course. Around the edges of the straightaway were cars, campers, kids and families in abundance. To try and land an airplane in the midst of all of this activity was certain disaster.

The 767 shown with emergency slides on each side of the fuselage. A portable generator is shown at the nose of the airplane. Photography by Maurice Quintal.
Pearson and Co-pilot Quintal turned toward Gimli and continued their steep glide. Flight 143 soon disappeared below Winnipeg's radar screens, the controllers frantically radioing for information about the number of "souls" on board. Approaching Gimli Pearson and Quintal made their next rude discovery: The RAT didn't supply hydraulic power to the 767's landing gear, flaps, or slats. Pearson ordered a "gravity drop" as Quintal thumbed frantically through the Quick Reference Handbook, or QRH. Quintal tossed the QRH aside and hit the button to release the gear door pins, hearing the main gear fall and lock in place. But he only got two green lights, not three. The nose gear, which fell forward against the wind, hadn't gone over center.

Six miles out Pearson began his final approach onto what was formerly RCAFB Gimli. Pearson says his attention was totally concentrated on the airspeed indicator from this point on. Approaching runway 32L he realized he was too high and too fast, and slowed to 180 knots. Lacking divebrakes, he did what any sailplane pilot would do: He crossed the controls and threw the 767 into a vicious sideslip. Slips are normally avoided on commercial flights because of the concern the buffeting creates in some of the passengers. Some of Flight 143's passengers ended up looking at nothing but blue sky, the others straight down at a golf course. Says Quintal, "It was an odd feeling. The left wing was down, so I was up compared to Bob. I sort of looked down at him, not sideways anymore.

Fuselage of the 767 straddling a guard rail. Photography by Maurice Quintal.
The only problem was that the slip had further slowed the RAT, costing Pearson precious hydraulic pressure. Would he be able to wrestle the 767's dipped wing back up before the plane struck the ground? Trees and golfers rushed by the starboard side passengers' windows as the 767 hurtled toward the threshold at 180 knots, 30-50 knots faster than normal. The RAT didn't supply "juice" to the 767's flaps or slats so the landing was going to be hot. Pearson didn't recover from the slip until the very last moment. A passenger on flight 143 reportedly said "Christ, I can almost see what clubs they are using." Co-pilot Quintal suspected Pearson hadn't seen the guardrail and multitude of people and cars on down the runway, but by this point it was too late to do anything about them. A glider only gets one chance at a landing. Quintal elected to stay silent.

The "Amazing Coincidences" of Gimli:

  • Pilot had extensive glider experience, co-owned a Blanik L-13
  • Co-Pilot had once been stationed at Gimli, was familiar with it.
  • Wpeg ATC had old style radar which allowed them to track 767 w/o transponder
  • Sports car club had fire extinguishers galore, Jaws of Life
  • ER Physician Colin Nesbit in Cessna doing preflight at end of 32R
  • RCMP Officer driving by airport saw 767 coming in, called for help
  • Air Canada mechanics driving a van to Gimli to begin repairs of #604 ran out of gas in the backroads of Manitoba.
Why did Pearson select 32L instead of 32R? Gimli was uncontrolled so Pearson had to rely on visual cues, and it was approaching dusk. Runway 32L was a bit wider, having been the primary runway at Gimli in prior years. Light stanchions still led up to 32L And the "X" painted on 32L, indicating its inactive status, was reportedly quite faded. Having made an initial decision to go for 32L, the wide separation of the runways would have made it impossible for Pearson to head for 32R at the last moment. Pearson says he "Never even saw 32R, focusing instead on airspeed, attitude, and his plane's relationship to the threshold of 32L."

The 767 silently leveled off and the main gear touched down as spectators, and kids on bicycles fled the runway. The gigantic Boeing 767 was about to become a 132 ton silver bulldozer. One member of the Winnipeg Sports Car Club was reportedly walking down the dragstrip, five gallon can full of hi-octane racing fuel in hand, when he looked up and saw the 767 headed right for him. Pearson stood on the brakes the instant the main gear touched down. An explosion rocked through the 767's cabin as two tires blew out. The nose gear hadn't locked down, and collapsed. The nose of the 767 slammed against the tarmac, bounced, and began throwing a hundred foot shower of sparks. The right engine nacelle struck the ground. The 767 reached the tail end of the dragstrip and the nose grazed a few of the guardrail's wooden support poles. (The dragstrip began in the middle of the runway with the guardrail extending out towards 32L's threshold.) Pearson applied extra right brake so the main gear would straddle the guardrail. Would all the sports car fans all be able to get out of the way, or would Pearson have to veer the big jet off the runway to avoid hitting stragglers?

"I've found myself that glider pilots make better jet drivers. The guys who do the best deadstick landings in the simulator are all glider pilots. It definitely gives you an edge in managing the flight."

- Jeff Morris, Training Captain with Hunting Cargo Airlines, Ireland

The 767 came to a stop on it's nose, mains, and right engine nacelle only a few hundred feet from spectators, barbecues and campers. The fuselage was intact. Inside, for an instant, there was silence. Then cheers and applause broke out among Flight 143's passengers. They'd made it; they were all still alive. But it wasn't over yet. Fire had broken out in the nose of the aircraft. Oily black smoke began to pour into the cockpit and front of the cabin. The fiery deaths of passengers in an Air Canada DC-9 that had made an emergency landing in Cincinnati a month before was on all the flight attendants' minds, and an emergency evacuation was ordered. The unusual nose-down angle the plane was resting at made the angle of some of the rear emergency slides nearly vertical. Descending them was going to be treacherous.

The only injuries that resulted from Pearson's dead-stick landing of Flight 143 came from passengers exiting the rear emergency slide too fast and hitting the asphalt. None of the injuries were life-threatening. The race fans had all managed to flee the path of the 767. The fire in the aircraft's nose area was battled by members of the Winnipeg Sports Car Club who reconverged on the plane with dozens of hand-held fire extinguishers. Pearson had touched down 800 feet from the threshold and used a mere 3000 feet of runway to stop. A general aviation pilot who viewed his landing described it as "impeccable." The 767 was relatively undamaged. Aircraft #604 was repaired sufficiently to be flown out of Gimli by Air Canada's chief CFI two days later, and after approximately $1M in repairs it re-entered the fleet. To this day Aircraft #604 is known to insiders as "The Gimli Glider." The avoidance of disaster was credited "To Capt. Pearson's knowledge of gliding, which he applied in an emergency situation to the landing of one of the most sophisticated aircraft ever built." Captain Pearson credits Co-pilot Quintal strongly for his cockpit management of "Everything but the actual flight controls," including his recommendation of Gimli as an alternate landing spot. Captains Pearson and Quintal spoke at the 1991 SSA Convention in Albuquerque about their experience. Pearson was, at the time, still employed and flying for Air Canada, and occasionally flying his Blanik L-13 on the weekends; he has since retired to raise horses. Maurice Quintal is now an A-320 Pilot for Air Canada, and will soon be captaining 767's - including Aircraft #604, "The Gimli Glider."
©1997WHN

An amusing side-note to the Gimli story is that after Flight 143 had landed safely, a group of Air Canada mechanics were dispatched to drive down and begin effecting repair. They piled into a van. They reportedly ran out of fuel en-route, finding themselves stranded somewhere in the backwoods of Manitoba.

Credits:

Special thanks to William and Marilyn Mona Hoffer, authors of "Freefall, A True Story" ISBN 0-312-92274-4, Boeing Aircraft 767 Chief Engineer Hank Queen, Mr. Len Gelfand, Colin Nisbet, Don Sigmundson, Don Kawal, Robert Pearson, Maurice Quintal, and others.

ABOUT THE AUTHOR

Wade H. Nelson is a freelance writer living in Durango, Colorado, which he calls "The last of the great ski towns." He may be reached at: wadenelson@frontier.net

Soaring Magazine 22 October 1997