COVER STORY The 156-tonne

GIMLI PHOTO : OREYACIE FMANITOBA OF ARCHIVES COURTESY

Grounded: Inspecting the damage after Flight 143’s unorthodox . It’s the 20th anniversary of aviation’s most famous .

piloting Flight 143 on a routine flight loss of pressure in the right main fuel Merran Williams from to Edmonton, via tank. Realising the situation was becom- N THE WESTERN side . The was lightly ing serious, Pearson quickly ordered a of Manitoba’s idyllic loaded, with 61 passengers and five crew. diversion to Airport, 120 miles Lake Winnipeg lies an Flight 143 climbed to its cruising alti- away. It became clear they were running old Royal Canadian Air tude of 41,000 feet and the first hour of out of fuel. Force station. As a town flight was straightforward for the experi- The left engine was the first to flame ofO just 2,000 people, Gimli is a tiny dot enced flight crew. However, just after out. At 2021, when their altitude was on the map, eclipsed by its larger neigh- 2000 local time, Pearson and Quintal 28,500 feet and they were 65 miles from bour, Winnipeg. But thanks to a 20-year- were shocked to see cockpit instruments Winnipeg, the right engine stopped. old accident, Gimli is probably the most warning of low fuel pressure in the left Flight 143 was gliding. Most of the famous landing ground in Canada. fuel pump. At first they thought it was a instrument panels went blank as they On July 23, 1983, Captain Bob Pearson fuel pump failure. had been relying on power generated by and First Officer Maurice Quintal were Seconds later, warning lights indicated the engines, and suddenly Pearson was

22 FLIGHT SAFETY AUSTRALIA JULY-AUGUST 2003 Sideslipping a 767 COVER STORY

Aircraft too high. Speed slowed to 180 knots flying blind. A magnetic compass, an out, the absence of a nose wheel saved artificial horizon, an airspeed indicator lives. The pilots were shocked to see peo- and an altimeter were the only instru- ple on the runway as they descended. ments still working. Unknown to Air Traffic Control, Gimli The dropped from airbase had become a two-lane dragstrip. near the right wheel well and used wind The rally spectators were startled to see power to turn a four-foot propeller, pro- a huge bearing down on them, viding enough hydraulic power to silent except for the rushing of wind manipulate the ailerons, elevators and against its body. People scattered as . However, the pilots were unable quickly as they could, but only the fric- to operate speed brakes, flaps or the tion between the aircraft nose and the Pilot initiates slideslip. Left undercarriage or carry out reverse thrust ground as the partly extended nosewheel rudder pedal on landing. collapsed, brought the aeroplane to rest pushed while he At 2031, realising Flight 143 did not in front of them. turns the yoke to have enough height to reach Winnipeg, The time was 2038 hours. Just 17 min- the right. Aircraft the pilots called Winnipeg Air Traffic utes had elapsed since Pearson had start- manoeuvred into a steep angle, Control to request a change in heading ed flying a powerless 767 from 28,500 Aircraft rapidly to Gimli, a decommissioned airforce feet to a safe landing. loses altitude. base 12 miles away. Gimli wasn’t listed Pearson and Quintal became in ’s manuals but, fortuitous- overnight celebrities and Gimli a house- ly, Quintal had been stationed there hold name across the world. An accident when serving in the airforce. As far as that came so close to tragedy ended as a anyone knew, both of its 6,800-foot run- triumph of human ingenuity. ways would be deserted. But while the crew of Flight 143 were As the aircraft descended without praised for their skill and bravery under power, Pearson needed all his flying skills pressure, a vital question remained. How to keep it on track. He had only one did an aircraft as advanced as a Boeing Aircraft chance to land – there could be no 767, with all its cutting edge avionic straightened up at altitude 40 missed approach. Unfortunately the air- technology, run out of fuel? feet. craft was coming in too fast and was A federal government public inquiry going to overrun the runway at its cur- carried out a comprehensive investiga- rent speed, as there was no way of apply- tion into the accident, using reports ing reverse thrust. compiled by Air Canada and the Pearson took a gamble that the 767 Transportation Safety Board of Canada would respond in the same way as a (TSB). Pearson himself was on the wit- smaller aircraft and executed a sideslip ness stand for five days and remembers by turning the yoke to the right at the seeing seven television cameras trained same time as he jammed his foot against on him amid the media frenzy on the the left rudder pedal. The aircraft first day. responded and descended enough to The reason for the accident turned out Aircraft touches down. bring it in on target. The manoeuvre to be all too familiar. Systemic problems required exceptional piloting skills as the with Air Canada training and proce- indicated airspeed wasn’t correct during dures, had led to a series of uncorrected the sideslip because the angle of the air- errors by ground and flight crew. The craft was different from its direction of TSB’s final report, a tome of almost 200 ILLUSTRATION travel. It came down to Pearson’s judge- pages, criticised Air Canada’s upper ment and experience as a glider pilot. management for serious communication Aircraft comes to During the nerve-wracking descent, failures. The TSB concluded that pro- : rest EE MARKMANN PETER Quintal tried using a back-up system to ducing manuals and procedures for per- lower and lock the landing-gear. The sonnel was a “corporate responsibility” gear on each wing was deployed but the not being adequately fulfilled by Air nosewheel stuck part way. As it turned Canada management.

FLIGHT SAFETY AUSTRALIA JULY-AUGUST 2003 23 GIMLI RED LAKE WINNIPEG

MONTREAL

OTTAWA

Cut short Travelling from Montreal to Edmonton via Ottowa, Flight 143 was forced to divert to Winnipeg because of a fuel shortage that led to the loss of both engines. When it became apparent the aircraft would not reach Winnipeg, the pilots changed course and headed to Gimli.

The flight and cabin crews were Although he couldn’t diagnose the praised for averting a major disaster “The failure of one exact problem, Yaremko found that if he through their “professionalism and skill” disabled the faulty circuit breaker, the which helped them overcome the prob- inductor coil should not backup circuit breaker got the gauges lems caused by “corporate and equip- working again and provided the required ment deficiencies”. have disabled the fuel fuel readings. The mechanic labelled the The trouble started almost three weeks pulled circuit breaker with yellow main- before the accident when the fuel quanti- gauges.” tenance tape to prevent it being turned ty indicating system on aircraft No. 604 back on. But he did not clearly record in (later Flight 143) was examined follow- the logbook his reasons for doing this. ing a directive from Boeing. As each fuel self-testing mechanism enabling it to The 767 flew from Edmonton to gauge was checked, it mysteriously went recognise faults within the system. Montreal via Ottawa without incident blank. However, a later check found the These built-in redundancies did not after the pilot in command satisfied him- gauges apparently working normally, so prevent the processor from failing, how- self that it was legal to operate the air- the aircraft was given clearance to fly. ever. Tests performed after the accident craft under provisions of the Minimum On the night of 22-23 July, the prob- found the failure was caused by a “cold Equipment List (MEL) despite the devia- lem resurfaced and the same mechanic, solder” joint on the inductor between tion reported in the fuel processor. Conrad Yaremko, investigated, unaware one coil wire and its terminal post. While Because of the unreliable electronic it was the same aircraft. He discovered a the terminal post was pretinned and had fuel monitoring system, when the air- malfunction in the digital fuel gauge enough solder sticking to it, the coil wire craft reached Dorval Airport in processor but was told no replacement end was not pretinned and had poor Montreal, maintenance worker Jean processors were available. adhesion. Ouellet was assigned to conduct a manu- The processor was a dual-channel sys- Still, the failure of one inductor coil al drip check of the aircraft’s fuel levels tem that provided fuel quantity meas- should not have disabled the fuel gauges. before its dispatch to Edmonton. He was urement, calculation and indication, and Another inductor coil in the second intrigued by the problem with the fuel was located under the aircraft’s floor, processor should have taken over if the processor and despite not having the behind the cockpit. It was considered the processor had performed according to its authority or training, took it upon him- “heart” of the fuel quantity indication specifications. Investigations revealed a self to tinker with the electronics while system on the Boeing 767 and was built design error was to blame. The processor waiting for the fuel truck. As he later told by Honeywell to Boeing specifications. failed to switch from the defective chan- investigators: “I thought I would do a Its benefits included an ability to operate nel to a working channel because there BITE [built-in test equipment] test on on a second channel if one failed, and a had been a drop in the power supply. the processor, so I pushed in the breaker

24 FLIGHT SAFETY AUSTRALIA JULY-AUGUST 2003 COVER STORY in the cockpit that was deactivated.” This nance workers performed a drip test and ing truck to tell when they had reached the made the fuel gauges blank again. estimated that 7,682 litres of fuel remained right number of litres to make up 22,300kg. The fuel truck arrived and Ouellet left in the tanks. The flight from Montreal to But the 767 was the first aircraft in Air the aircraft without deactivating the Edmonton, including a brief stop in Canada’s fleet to use metric units (kilo- faulty circuit breaker. As the investigation Ottawa, required 22,300kg of fuel, an grams) rather than imperial (pounds). later reported, “the well-intentioned but amount expressed as mass because of the Metric units were being phased in across misguided curiosity of Mr Ouellet result- importance of knowing an aircraft’s weight. Canada, and the conversions were still ed in blank fuel gauges in the cockpit, The mechanics needed to work out how causing confusion. and contributed significantly to the sub- many litres made up 22,300kg. They could With the help of First Officer Quintal, sequent accident.” then subtract the 7,682 litres already in the the ground crew used the correct proce- With the fuel gauges inoperative, mainte- tanks, and use the fuel gauge on the refuel- dure to calculate the weight in kilos.

OUT OF FUEL AT 28,500 FEET: A NIGHTMARE SCENARIO

YOU’RE FLYING a Boeing 767. Both engines fail ably would have made it easier for us.” way was to sideslip the giant aircraft on the final and you’re left to glide at 28,500 feet. What do A few months after the incident, Pearson approach so it would touch down close enough you do? For Air Canada captain Bob Pearson, learnt that Scandinavian Airlines had made it to the beginning of the runway that it wouldn’t there was only one answer: keep flying the mandatory for pilots to perform a successful run out of tarmac. This manoeuvre was unprece- aircraft using all the techniques you’ve ever deadstick landing through the simulator before dented. Fortunately, it worked, and Flight 143 learnt. they were endorsed to operate any new type of touched down safely. On July 23, 1983, Pearson and his first officer aircraft. But Pearson is relieved that he wasn’t flying Maurice Quintal faced a nightmare scenario While Pearson is modest about the piloting an Airbus. “You can’t sideslip an Airbus aircraft, when a fueling mishap left them powerless on skills he used to bring Flight 143 to a safe the computers won’t let you,” he says. “Boeing what was expected to be a routine flight from landing, his experience as a gliding and aero- aircraft are capable because they’re a hydraulic- Montreal to Edmonton. batic instructor was essential when it became controlled aircraft and you can cross control.” Twenty years later, Pearson, now 68, is apparent that the aircraft was travelling too fast The Gimli accident led to a huge investigation matter-of-fact about the incident that made him to land on the runway at Gimli airbase near by both the Canadian Transport Safety Board and a household name in aviation. He notes that all Winnipeg. the Federal Government Public Board of Inquiry. pilots should know how to glide. “The pilot Pearson needed to lose altitude fast. The only Air Canada’s own investigation blamed the pilots closes the throttle so they end up at flight idle and mechanics for the accident. The official setting rather than ground idle which is slower. investigations cast the net wider and concluded At that power setting, the power coming out of that Air Canada’s procedures were at fault. the back of the engine offsets the drag of the Pearson spoke out against the Air Canada front so it’s as if the engine isn’t there. Every verdict at the time and says he hasn’t stopped pilot is actually gliding on normal descent [if the talking to the media. “I don’t think any other power levers are closed].” The difference with employee of Air Canada has ever done that,” he the “”, as the incident came to be laughs. “I was not afraid for my job even though known, is that no one had ever successfully I was working for them – I think they were a little landed a 767 without power. bit afraid of me!” Pearson’s Canadian pilot’s licence required Pearson flew for Air Canada for the next 10 him to demonstrate the ability to perform a years, retiring at 58. He then flew for Asiana deadstick landing in a single-engine aeroplane, Airlines until retiring eight years ago, aged 60. but he and Quintal had no way of knowing The Gimli Glider incident received worldwide whether the techniques would work in a 767. attention and became the subject of a best “I don’t think anyone anticipated this sort of selling book and a TV movie. Pearson had a bit

thing happening,” Pearson says. “On a twin- PHOTO part in the movie, Falling from the Sky: Flight engine aeroplane, you’re trained ad nauseum to 174: you can see him holding a clipboard when : fly on one and on a four-engine aeroplane you’re PEARSON BOB COURTESY the pilots come out of the simulator. trained to land on two. But at Air Canada, we had However, he was unimpressed by the movie’s no training on what to do if both engines fail.” accuracy and was frustrated to see mistakes in Pearson sees this as a failing in pilot training. aviation terminology. He preferred the book “Year after year, pilots attend training and learn Freefall, by Marilyn and William Hoffer. the same stuff. My feeling is that pilots should Pearson now travels extensively as a public get more of these unusual situations. If we had speaker, and is in high demand as an expert practised in the simulator, even once, it prob- Bob Pearson at the time of the Gimli accident aviation witness in legal proceedings.

FLIGHT SAFETY AUSTRALIA JULY-AUGUST 2003 25 Gimli slider The only injuries that came out of Flight 143’s powerless landing occurred when passengers used the emergency slides to evacuate the aircraft after it came to rest at Gimli Airbase.

However, they had not been trained in cor- would give them the correct volume in Using a computer to calculate fuel also rect conversion, so the figure of 1.76 pro- litres. Through this calculation, the crew caused confusion over responsibilities. In vided by the refueling company on their determined that 4,916 litres needed to be the past, when fuel was calculated manu- refueling document, was taken to be the added from the fuel truck. The problem ally, a flight engineer’s duties included required multiplier. It was typical of the was that 1.77 is the multiplier that con- checking the fuel load. Flight engineers numbers seen on previous slips and they verts litres into pounds, not kilograms: to were a thing of the past on this 767, as a assumed that the numbers provided over convert litres into kilograms you need to Presidential task force, under Ronald the previous few months had indicated multiply by 0.8. Flight 143 did not have Reagan, had determined that aircraft specific gravity in the new metric system. 22,300kg on board, it had about could be built to be operated by two They decided to multiply 7,682 by 10,000kg, less than half the amount of A- pilots instead of three, if the tasks previ- 1.76. This would mean 13,597kg 1 kerosene jet fuel needed to get the air- ously given to the second officer (flight remained in the tanks, requiring an infu- craft to Edmonton. The refueller didn't engineer) were either fully automated or sion of 8,703kg to bring the fuel level up know where the flight was headed, so no handled by ground staff. Responsibility to 22,300kg. They then divided 8,703kg alarm bells rang for him as he poured for ensuring adequate fuelling had by 1.76, assuming that this conversion fuel into the tanks. passed to the maintenance branch. But

26 FLIGHT SAFETY AUSTRALIA JULY-AUGUST 2003 COVER STORY

“There had already been 55 changes to the MEL in the three months Air Canada had been operating the 767.”

characteristics of the 767. They did not had become such a fluid document, the appear to have sufficient background perception grew that maintenance per- knowledge of these aspects of the air- sonnel, rather than the MEL, should have craft. They did not appear to have the final word on the aircraft’s airworthi- received sufficient training about some ness. And maintenance control cleared of the critical aircraft systems, in particu- Flight 143 for flight. lar, the fueling system.” The decision was further clouded by the The investigation also revealed an number of people who crowded into the organisational shortcoming at the Air cockpit just before departure. Pearson, Canada maintenance base. The crew held Quintal and flight attendant Anne Swift morning meetings at which senior techni- later testified that between six and eight cal experts gathered to discuss the major people visited the front of the aircraft after maintenance issues for each aircraft in the the flight crew had boarded. They includ- fleet. However, the meetings took place ed maintenance personnel, the fueller, an only five mornings a week: from Monday Air Canada pilot travelling as a passenger, to Friday. July 23, 1983 was a Saturday. and flight attendants. The TSB report Despite being unaware of the inade- concluded that these people could have quate fuel load, Pearson took the absence distracted the flight crew at a crucial time of computerised fuel monitoring seri- in their departure preparations. ously and carefully consulted the 767 Pearson then made the fateful decision minimum equipment list (MEL). The

PHOTO to allow the flight to proceed. He took off MEL stated that one main fuel tank at full throttle, which lifted the 767 : OREYACIE FMANITOBA OF ARCHIVES COURTESY gauge could be inoperative when the air- quickly into the air and disguised the air- craft was dispatched. However, Air craft’s lightness, a condition that might Canada maintenance told the pilots that have caused the pilots to question it was legal to operate with both the main whether they had their full load of fuel. tank gauges unserviceable, as indicated What happened next soon became his- by the master MEL (provided by Boeing) tory and still stands as a cautionary tale as long as a full drip was conducted on to airlines, pilots and maintenance per- the aircraft’s fuel tanks. This was backed sonnel alike. up by a page in the Boeing operating The tale has an interesting postscript. because these men were not trained to manual, which was removed shortly after After putting the aircraft down, Pearson calculate fuel, they assumed the pilots the accident, and by prompts on the pro- was left waiting at Gimli until two in the would make sure it was done properly. gramable management computer. At no morning for the Air Canada mechanics The problem was neither of the pilots time did the pilots believe they were not dispatched to assess the damage. Driving was trained in this technical task. Safety operating legally. through the back roads from Winnipeg, procedures had failed to keep pace with Pearson also had to consider that if he they had run out of fuel. new technology. As the investigation later grounded too many flights, it would concluded: “Air Canada ... neglected to reflect on his professional abilities. He had Acknowledgments: Final Report of the Board assign clearly and specifically the respon- to be absolutely sure there was no way the of Inquiry into Air Canada Boeing 767 C- sibility for calculating the fuel load in an flight could proceed. This 767 aircraft was GAUN Accident – Gimli, Manitoba July 23, abnormal situation.” so new its MEL had some blank pages 1983; William and Marilyn Hoffer: Freefall: The investigation attacked Air because the procedures were still being From 41,000 feet to zero – a true story; Wade H. Canada’s training procedures, noting developed. Nelson: “The Gimli Glider”, Soaring: The Journal “both flight crew and maintenance per- There had already been 55 changes to of the Soaring Society of America, October 1997; sonnel seemed to be somewhat mes- the MEL in the three months Air Canada William Carley, “Out of Fuel at 26,000 Feet”, Wall merised by the complex, computerised had been operating the 767. The MEL Street Journal, December 4, 1984.

FLIGHT SAFETY AUSTRALIA JULY-AUGUST 2003 27