The crashes of Air Ontario flight 1363 and USAir flight 405
On the 10th of March 1989, Air Ontario flight 1363 failed to become airborne on takeoff from Dryden, Ontario and crashed into a forest, killing 24 of the 69 people on board. The investigation by Canada’s transportation safety board uncovered a tragic confluence of events that led to the Fokker F-28 taking off with ice on its wings. In the process, it discovered massive deficiencies in the way pilots, airlines, and airports treated the problem of wing contamination. Then, on the 22nd of March 1992, USAir flight 405, another Fokker F-28, failed to get into the air as it tried to take off from New York’s LaGuardia Airport. The plane crashed off the runway and into Flushing Bay, killing 27 of the 51 people on board. After this second accident, the Federal Aviation Administration launched an overhaul of the way planes were de-iced — one that the investigation into the Air Ontario crash had already recommended.
Flight 1363 was a regularly scheduled flight with the regional carrier Air Ontario from Thunder Bay, Ontario to Winnipeg, Manitoba, with a stopover in the remote town of Dryden. The flight was operated by a Fokker F28 ‘Fellowship,’ a Dutch short-range jet with two rear-mounted engines and capacity for 65 passengers. Air Ontario had just acquired two F28s in 1987 and the crew that would be flying flight 1363 did not have much experience with the type. Although both pilots were very experienced and were familiar with flying in remote parts of Canada, Captain George Morwood had only been flying the Fokker F28 for two months, and First Officer Keith Mills had flown the F28 for only one month.
Air Ontario was a so-called “feeder airline” to Air Canada, which owned a 75% stake in the company. However, a large portion of Air Ontario’s staff were holdovers from a recent merger with the bush operator Austin Airways, which conducted flights around the Hudson Bay to airports with limited or no services using very small aircraft. Internal chaos ensued at the airline as disagreements between line pilots and bush pilots erupted into strikes and management gridlock. It was in this environment that Air Ontario acquired the two Fokker F28 Fellowship aircraft, the first jets ever flown by the airline. Air Ontario proved unable to retain experienced personnel who could help the carrier adjust to jet operations and instead hired pilots with virtually no jet experience. Some of the new hires never got any simulator time, and the check pilots training them often had only a little bit more experience than their students. By March 1988, the airline still lacked operations manuals and minimum equipment lists for the F28s, leaving pilots in the dark about the finer performance limits of the aircraft, what types of problems were supposed to preclude takeoff, and what regulations applied to the model. Furthermore, some pilots who received F28 training with other airlines stuck to those airlines’ procedures, even though they sometimes conflicted with one another. Air Canada exercised a hands-off approach to its majority stake and there was no sign that anyone at Canada’s flag carrier knew about the massive operational deficiencies at Air Ontario.
It was in this environment that Captain Morwood and First Officer Mills prepared for the flight from Thunder Bay to Winnipeg via Dryden on the 10th of March 1989. Weather in the region was poor that day, with large areas of snowfall and temperatures between 0 and -1˚C. While on the ground in Thunder Bay, Captain Morwood was informed by an Air Ontario dispatcher that Air Canada had cancelled a flight, and that 10 passengers from that flight would be placed on his plane instead. This was on top of 55 passengers and four crew already scheduled to board, meaning the plane would be at maximum capacity. Morwood’s calculations showed that with these extra passengers and their bags, the plane would be over its maximum takeoff weight. He wanted to remove some of the passengers, but was told to offload fuel instead. This meant that he would have to take on more fuel during the stop in Dryden, Ontario.
Flight 1363 finally departed Thunder Bay at 11:55, an hour late, after adding the new passengers and offloading 1,280kg (2,822lb) of fuel. Further complicating the situation was that the plane’s Auxiliary Power Unit (APU) was not working. This electrical generator is used to power the aircraft when the engines are turned off, and to start the engines before taxiing. The small airport in Dryden lacked the appropriate equipment to jump start the engines, so if the engines were shut off during the stopover, it would be impossible to restart them and the plane would be stuck there indefinitely, causing the cancellation of the flight. This meant that the pilots would have to perform a “hot refueling” with one engine running, a practice which was known to be dangerous. In fact, the minimum equipment list stated that the APU must be functional in order to take off, but neither the pilots nor the dispatcher had access to the minimum equipment list for the F28 and they did not know this.
After landing in Dryden 45 minutes later, Captain Morwood refueled the plane, and was then observed having an angry conversation with Air Ontario managers about the situation. Growing self-conscious about the mounting delays, and observing that the weather was deteriorating toward minimums, Captain Morwood apologized to his passengers and prepared to leave Dryden. The pilots elected not to de-ice the plane, even though snow was actively falling. This was because it was forbidden to de-ice the plane with an engine running, which could cause antifreeze fumes to enter the cabin. And if they shut off the engines, they wouldn’t be able to start them again, so the flight would have to be cancelled and passengers rebooked. Besides, during their bush flying experience, ice was never a big problem. But a Fokker F28 Fellowship is not a bush plane. Had the pilots been familiar with the operations manual, they would have known that it was forbidden to take off with any ice on the wings whatsoever, because the F28 had much lower aerodynamic margins than most other airliners. Even a millimeter of ice on the wings could disrupt airflow and cause a 50% reduction in lift.
As flight 1363 sped down the runway, it seemed sluggish and took longer than usual to lift off the ground. And even after rotating, the main wheels stayed on the ground for some time, until the pilots added more power and pulled back further. As soon as the plane got into the air, it was clear it wouldn’t stay there for long. It started rolling from side to side, its wings almost scraping the ground, and overflew the end of the runway at an altitude of just 15 feet. At the end of the runway was a valley and the plane immediately started descending into it, striking the tops of trees with the landing gear and wingtips. Flight attendants shouted for passengers to assume the brace position. Just a few hundred meters beyond the end of the runway, the F28 plowed headlong into another grove of trees, ripping off the left wing. The plane landed sideways on the snow-covered ground, breaking into three main sections and exploding in flames.
The impact instantly killed many people on the left front part of the plane, including both pilots, but the majority had survived and were faced with a desperate rush to escape. Many were seriously injured after seats tore out of their mountings and piled against each other. The exits on the left side were blocked by fire, so most people escaped through breaks on the right side of the fuselage. After making it safely outside with his family, one man returned to the plane and pulled twelve more people out of the burning wreckage. Other passengers worked to free those who were trapped, while the smoke and heat grew more and more intense. But once everyone was out of the plane, some of them having sustained serious burns in the process, they faced hypothermia as they waited for rescue in the freezing forest. By the time rescue crews had finished scouring the crash site, it became clear that 22 people were dead while 47 survived. Of those, two more died in hospital, bringing the final death toll to 24.
The commission of inquiry into the crash, led by the Honourable Virgil Moshansky, found a large array of factors that all led to the decision not to de-ice the plane before takeoff. First, there was the fact that the plane was not allowed to de-ice with the engines running. This was the fault of the airline, which should never have dispatched the plane with an inoperative APU, but lacked the minimum equipment list that would have told them this. Second, both pilots were under pressure to leave. Morwood was known for his attention to passenger comfort and was frustrated that the flight was delayed. He and Mills also had plans the next day; cancelling the flight would be sure to sink them. And third, it was clear that neither pilot understood the danger that ice posed to the F28. If they had known that ice could so easily cause a crash, they might have elected to cancel the flight. This represented a major regulatory oversight: how was it possible that two pilots who apparently knew so little about the capabilities of their aircraft in icing conditions ended up flying it in a harsh Canadian winter? And indeed, how was it possible that the company could go so long without a minimum equipment list, and allow a plane to fly with an inoperative APU, a malfunction that should have grounded it?
In addition, there were numerous flight attendants, pilots riding as passengers, and even regular passengers who saw the ice on the wings but didn’t tell Captain Morwood and First Officer Mills. Most believed that the plane would be de-iced and only realized that it would not shortly before takeoff. Furthermore, the flight attendants knew that Air Ontario usually treated their input as meaningless. And the off duty pilots considered it impolite to point out safety concerns to other pilots, who are assumed to know what they’re doing. It was evident that a corporate airline culture which did not value the input of anyone other than the pilots flying had played a role in the crash.
The commission’s final report found that all these small shortcomings stemmed from Air Ontario’s complete failure to comply with organizational regulations, lack of crew resource management training, and poor corporate culture. These factors were not identified because Transport Canada lacked enough properly trained staff to carry out its oversight mandate, and in fact the agency had been warned of this multiple times in the years leading up to the crash. In the end, the commission issued a number of sweeping recommendations, including an overhaul of training surrounding the dangers of wing contamination, and called for an increase in Transport Canada’s staff and oversight capabilities. Over 100 other recommendations were made to address the many unsafe practices and regulatory shortcomings uncovered during the investigation, not all of which directly contributed to the crash. One of these recommendations was that so-called “Type 1” de-icing fluid be phased out. Type 1 de-icing fluid is runny and is applied hot to the wings, removing the ice immediately but losing its effect after as little as six minutes. The commission preferred the use of Type 2 de-icing fluid, a gel that is applied cold, removing the ice and preventing the formation of new ice for up to 45 minutes. This particular recommendation would soon be the focus of much interest.
Fast forward to the 22nd of March 1992, well after the initial publication of the recommendations put forth by the Moshansky commission. On this day, another Fokker F28 Fellowship operating USAir flight 405 from New York to Cleveland was preparing to take off from LaGuardia Airport. The flight was already over an hour late when it arrived at LaGuardia, and further delays were adding up quickly. First, the pilots elected to de-ice the plane at the gate using type 1 de-icing fluid, which was still the mostly widely used type. But after the de-icing, a further 20-minute delay occurred because one of the de-icing vehicles broke down behind the plane and prevented it from taxiing away from the gate. By the time it was fixed, the de-icing fluid had lost its punch, and the pilots elected to de-ice the plane again. Finally, the plane left the gate at 21:00 — now an hour and 40 minutes late — with 47 passengers and four crew on board.
But the flight was soon delayed again. One of LaGuardia’s two runways was temporarily closed so it could be sanded, resulting in a long queue of aircraft waiting to take off on the one remaining runway. Over the next 35 minutes, flight 405 sat in the queue as very light snow fell on the airport. During this time the pilots certainly thought about ice — in fact, First Officer John Rachuba repeatedly turned on lights on the wings so he could look back and check for ice contamination. He apparently saw none, remarking to Captain William Majure, “Looks pretty good to me, from what I can see.” Even so, if they had wanted to de-ice the plane again, they would have lost their spot in line — and this might have put them back in the exact same predicament later, if it didn’t cause the flight to be cancelled altogether.
As it turned out, ice was indeed forming on the wings as the effectiveness of the type 1 de-icing fluid quickly wore off. But neither pilot could see the ice because the amount that had formed, while certainly dangerous, was not visible from the cockpit, even though USAir crews universally believed that it would be. Flight 405 was finally cleared to take off at 21:35 with its pilots totally unaware that the ice on the wings was considerably increasing their stall speed. Captain Majure decided on a lower V1 speed than usual (that is, the speed above which the takeoff cannot be aborted) because of the possibility of slush on the runway. This would have an unintended side effect. On the Fokker F28, V1 and VR (the speed at which the nose is rotated upward) are normally the same, but with a lower V1, they were now different. However, as the plane accelerated down the runway, First Officer Rachuba instinctively called out VR immediately after V1, leading Captain Majure to rotate prematurely.
The ice contamination was already reducing the capability of the wings to generate lift, and the early rotation may well have been the final straw that prevented the plane from getting off the ground. Both the ice and the early rotation led to an increased angle of attack — the angle of the nose relative to the airstream — and subsequently a stall. Flight 405 floated along just a couple feet above the ground, unable to find the lift to climb. The pilots immediately realized that their plane would not fly, but there was little they could do to prevent a crash. The left wing hit the runway, sending up sparks and dragging the plane to the left onto the grass. It hit several indicator posts, touched down briefly, bounced back into the air, struck the ILS localizer beacon, and demolished a pump house, which ripped off the left wing. Breaking apart as it went, flight 405 rolled over the seawall and crashed inverted into the shallow waters of Flushing Bay.
The crash killed 12 people outright, but the remaining 39 were now faced with the simultaneous threats of burning and drowning. Passengers and crew in the front of the plane found themselves hanging upside down with their heads underwater. The rest of the plane had come to rest upright half-submerged in the bay, but much of what was above the surface quickly caught fire. Passengers fumbled to undo their seat belts and escape through the freezing water. Some clambered up the seawall and staggered onto the runway, while others clung to floating wreckage and were picked up by firefighters who arrived on the scene almost immediately. Many more never made it out. In addition to the 12 killed on impact, 15 drowned after the crash, bringing the death toll to 27 while 24 survived. Among the dead was Captain Majure, but First Officer Rachuba managed to escape.
Investigators found that USAir pilots had been taught about the dangers of icing, but they were not taught effective ways to detect it. The simple fact is that wing contamination cannot be reliably seen from the cockpit on any airliner. Procedures called for pilots to look from the cabin if they were unsure, but most pilots believed the view from the cockpit was just as good. In fact, the only way to be certain whether there is ice on the wings is to physically touch them. But pilots everywhere were taking off with ice on their wings because it was often impossible to de-ice the plane immediately before takeoff so that the type 1 de-icing fluid would be at full strength.
This represented a major problem in the airline industry — one which could have been addressed earlier. Better training around the danger of ice and a stronger de-icing substance were both recommendations that came out of the crash of Air Ontario flight 1363 that could have prevented the crash at LaGuardia. USAir had trained its pilots in the danger of ice but did not provide ways for pilots to know with certainty whether their plane had ice on it. When it came to a decision between taking off with the possibility of ice, when none had been seen, or cancelling the flight, pilots were understandably reluctant to cancel the flight. And flight 405, like every other plane at LaGuardia that night, was de-iced using type 1 de-icing fluid, which was known to be ineffective. The Moshansky Commission’s interim report, including these recommendations, had been published in 1989, only a few months after the accident in Dryden, but somehow the FAA had not considered its recommendations and the Air Ontario crash was not referenced in the NTSB report on USAir flight 405!
It remains unclear to this day why no one at the FAA knew about Moshansky’s findings. Years later, Moshansky asserted that he sent the interim report to the FAA, but that it probably ended up “sitting on a desk somewhere” and never got to the right people. The commission’s findings presumably would have circulated in industry publications, but at USAir, the rapidly growing airline lacked established communication avenues to get such information out to pilots like Majure and Rachuba who had only recently joined USAir with its acquisition of other airlines like Piedmont and Empire. The result was that the lessons from the crash of Air Ontario flight 1363 not only didn’t make it to the pilots of USAir flight 405, but in fact never made it out of Canada at all.
After the USAir 405 crash, the NTSB recommended many of the same things that Moshansky had recommended years earlier, and the FAA finally took action. Today, all pilots are trained to treat wing contamination with the utmost caution, especially on vulnerable aircraft like the Fokker F28. Type 1 de-icing fluid is now only used to clear snow, and if icing conditions exist it is always followed by type 4, which can prevent the formation of ice for up to two hours after application. And another recommendation from the Moshansky report, that de-icing facilities be placed near the runway so planes could de-ice right before takeoff, is also now widely implemented. (It is worth noting that the FAA dragged its feet on this recommendation because the installation of equipment near the runway posed a hazard in runway runoff scenarios. This view was eventually dropped.)
The lessons of these two crashes are very far reaching. Not only did they help revolutionize the industry’s handling of wing contamination, they also serve as a harsh reminder of the importance of communication. If communication between the commission of inquiry in Canada and the FAA in the United States had been more standardized, the Moshansky report would not have slipped through the cracks, and 27 people might not have died on USAir flight 405. Today, it is highly unlikely that the FAA would ever again overlook a report on a major accident — thanks in part to the vastly more interconnected world in which we now live. And finally, this pair of crashes underscores the fundamental driving principle behind why we investigate aircraft accidents: that change must come from every crash, lest we run the risk of letting it happen again.
_______________________________
Visit reddit.com/r/admiralcloudberg to read 100+ similar articles. This article was originally posted to Reddit on 29 June 2019.
